University of Ghana http://ugspace.ug.edu.gh UNIVERSITY OF GHANA COLLEGE OF BASIC AND APPLIED SCIENCES OPTIMIZATION AND CHARACTERISATION OF A 3-BLEND PLANT MILK BEVERAGE AMA FREMPOMAA ODURO DEPARTMENT OF NUTRITION AND FOOD SCIENCE JULY, 2018 University of Ghana http://ugspace.ug.edu.gh UNIVERSITY OF GHANA COLLEGE OF BASIC AND APPLIED SCIENCES OPTIMIZATION AND CHARACTERISATION OF A 3-BLEND PLANT MILK BEVERAGE BY AMA FREMPOMAA ODURO (10069752) THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE AWARD OF MPHIL FOOD SCIENCE DEGREE DEPARTMENT OF NUTRITION AND FOOD SCIENCE JULY, 2018 University of Ghana http://ugspace.ug.edu.gh DECLARATION I declare that this work presented as a dissertation to the Nutrition and Food Science Department, Legon, was carried out entirely by me under the supervision and guidance of Dr. Maame Yaakwaah Blay Adjei and Professor Firibu Kwesi Saalia of the University of Ghana, Legon. With the exception of references to cited works and textbooks which have been duly acknowledged, this dissertation has not been presented either in part or whole elsewhere for any other degree. Ama Frempomaa Oduro ……………………… ……………………… (10069752) Signature Date Certified by supervisors Dr. Maame Yaakwaah Blay Adjei ……………………. ……………………… Signature Date Prof. Firibu Kwesi Saalia ……………………… ……………………… Signature Date i University of Ghana http://ugspace.ug.edu.gh ABSTRACT Plant milk beverages have become a necessity as they serve as an alternative for people with lactose intolerance, cow milk allergy, vegetarians and people who want to avoid cholesterol. They have numerous health benefits but their consumption is limited by their poor sensory profiles. Soymilk is the most popular and most widely consumed plant milk beverage in the world. Using local ingredients for commercial production of plant milk in Ghana is advantageous as these raw materials are easily available and are affordable. Preliminary work used melon seeds, peanuts and coconut to formulate one, two and three blend plant milk beverages, but their sensory profiles suggested that they will be unappealing to consumers. The purpose of this work was to use a consumer led approach to optimize and characterize those plant milk beverages. A focus group discussion using target consumers was organized to confirm the selection of the three ingredients (melon seeds, peanuts and coconut) used for the initial plant milk beverages and to understand which sensory properties of plant milk are desirable to consumers. The plant milk beverages were optimized based on the outcome from the focus group discussions. After optimization, the 3-blend beverages were then characterized by sensory methods (consumer acceptability tests using Relative Preference Mapping (RPM) and the traditional 9-point hedonic scale in a Balanced Incomplete Block design (BIB) and physicochemical analysis (proximate composition, pH, titratable acidity, color, and viscosity). RPM was used to test liking for the optimized beverages as well as compare them to existing commercial dairy milk and plant milk beverages. Consumer test using BIB was used to understand the effect of the various ingredients on consumer acceptability. During the focus group discussions, only coconut met consumer expectations, peanuts was to be roasted instead of blanched and tiger nuts was recommended as a suitable ingredient to be added to the other ingredients. From the RPM, two of the 3-blend ii University of Ghana http://ugspace.ug.edu.gh beverages containing peanuts, tiger nuts and coconuts milk had liking scores comparable to the dairy milk and commercial plant milk beverage. The results from the consumer test using BIB showed that peanuts and coconut milk were responsible for the acceptability of the blends. Melon seeds milk decreased liking and tiger nuts milk had a fairly neutral effect on liking. The protein content of the reformulated 3-blend beverages was in the range of 1.65±0.06%- 3.51±0.02%, the fat content was from 4.09±0.05%-6.54±0.04%, the ash content was from 0.25±0.01%-0.51±0.07%, carbohydrates from 4.28±0.01%-7.26±0.01%, the L values were in the range of 75.91±0.13-85.54±0.50, the range for pH was 6.74±0.00-6.87±0.01 and that for titratable acidity was 0.83±0.01-2.05±0.00. The viscosity of the beverages was from 1163.55±7.42-1814.85±2.47 centipoise. Using consumer insight led to optimization of the plant milk beverages leading to acceptability comparable to an already accepted commercial plant beverage. iii University of Ghana http://ugspace.ug.edu.gh DEDICATION I dedicate this this work to God, Almighty, who made all this possible; to my late father Mr. Kofi Amoah Oduro, parents and siblings for their support, love and encouragement throughout this entire journey and to Nkunim, my nephew, who always reminded me to do my homework. iv University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT This project was made possible with financial support from the University of Ghana Research Fund. I am also grateful to my supervisors Dr. Maame Yaakwaah Blay Adjei and Professor Firibu Kwesi Saalia, throughout this research you were always ready to offer advice and guidance to make this work successful. I am thankful for all the time you spent to help me at various stages of my work. You were so patient and always ready to make time to discuss this work. I am most grateful to Mr. Asamoah and Auntie Joyce Duah and all the lab technicians who helped me with my processing and analysis. I am also grateful to staff of the Sensory Evaluation Laboratory: Sylvia, Elma, Cecille and Priscilla. Thank you for your help, encouragement and advice. To Salsabila Osman, my Mphil colleague (you were a co-researcher in this work and a companion on this journey). God bless you for your help with my processing, your advice and insight that made this work successful. I also want to thank Winifred Adu-Hanson, Claudia Sam, Henrietta Adu-Hanson and Lydia Mensah. Thank you so much for your help with my processing and data collection. Winifred, you were my chief recruiter! God bless you. I am so grateful to my dad, Mr. Clement Kwei. Thank you for your encouragement, support and for editing my work. I am grateful to my mother Mrs. Ruby Kwei, who prayed for me and kept vigil with me. Thank you, Maame and Alex and my siblings for your support through my master’s education. Thank you, Anthony for your encouragement and support. God bless you all!!!! Finally, every thanks goes to God for seeing me through my master’s education. This work has been possible only with His help and guidance. v University of Ghana http://ugspace.ug.edu.gh ETHICS The sensory aspects of this work were organized under the auspices of the Sensory Evaluation Laboratory of the University of Ghana which has ethics approval to carry out sensory evaluation tests into various foods at the Department of Nutrition and Food Science. vi University of Ghana http://ugspace.ug.edu.gh Table of Contents DECLARATION ........................................................................................................................................... i ABSTRACT .................................................................................................................................................. ii DEDICATION ............................................................................................................................................. iv ACKNOWLEDGEMENT ............................................................................................................................ v ETHICS ....................................................................................................................................................... vi LIST OF FIGURES ..................................................................................................................................... xi LIST OF TABLES ......................................................................................................................................xiii CHAPTER ONE ........................................................................................................................................... 1 1.0 INTRODUCTION .................................................................................................................................. 1 1.1 Rationale ............................................................................................................................................. 3 1.2 Main Objective .................................................................................................................................... 4 1.3 Specific Objectives ................................................................................................................... 4 CHAPTER TWO .......................................................................................................................................... 5 2.0 LITERATURE REVIEW ....................................................................................................................... 5 2.1 Consumer led product development ................................................................................................... 5 2.1.1 Stages in product development .................................................................................................... 5 2.1.2 Consumer research methods in product development ................................................................. 6 2.2 Product Optimization and Characterization ........................................................................................ 7 2.3 Emerging consumer trends leading to plant milk product development ........................................... 11 2.4 Plant Milk.......................................................................................................................................... 12 2.4.1 Preparation of Plant milk ........................................................................................................... 13 2.4.2 Advantages of consuming plant milk ......................................................................................... 14 2.4.3 Disadvantages of Plant milk ...................................................................................................... 15 2.4.2 The use of local plants in the commercial processing of plant milk in Ghana ........................... 17 2.5 Preliminary work on the initial plant milk beverages ....................................................................... 20 CHAPTER THREE .................................................................................................................................... 26 3.0 METHODOLOGY ............................................................................................................................... 26 3.1 Study design ...................................................................................................................................... 26 vii University of Ghana http://ugspace.ug.edu.gh 3.2 Methods............................................................................................................................................. 28 3.3. Focus group discussions .................................................................................................................. 28 3.3.1 Participants ................................................................................................................................. 28 3.3.2 Organization of the focus group ................................................................................................. 29 3.3.3 Products used for focus group .................................................................................................... 30 3.4 Plant milk optimization ..................................................................................................................... 31 3.4.1 Raw materials used for plant milk processing ........................................................................... 32 3.4.2 Plant milk processing ................................................................................................................. 33 3.5 Characterization of optimized beverages .......................................................................................... 36 3.6 Physicochemical characterization ..................................................................................................... 36 3.6.1 pH and Titratable Acidity .......................................................................................................... 36 3.6.2 Color .......................................................................................................................................... 37 3.6.3 Viscosity .................................................................................................................................... 37 3.6.4 Proximate composition .............................................................................................................. 37 3.7 Sensory characterization of optimized products ............................................................................... 38 3.8 Comparing the sensory acceptability of 3-blend samples with commercial dairy and plant milk beverages using Relative Preference Mapping (RPM) ........................................................................... 38 3.8.1 Test locations ............................................................................................................................. 38 3.8.1 Participants ................................................................................................................................. 39 3.8.3 Products ...................................................................................................................................... 40 3.8.4 General test procedure ............................................................................................................... 41 3.9 Consumer acceptance test of all 19 formulation prototypes using the BIB design to determine which of the main ingredients in the 3-blend plant milk prototypes influenced consumer liking .......... 43 3.9.1 Products ...................................................................................................................................... 43 3.9.2 Test Location ............................................................................................................................. 43 3.9.3 Participants ................................................................................................................................. 43 3.9.4 General Test procedure .............................................................................................................. 44 3.10 Data Analysis .................................................................................................................................. 45 3.10.1 Focus group discussion ............................................................................................................ 45 3.10.2 Physicochemical characteristics ............................................................................................... 45 3.10.3 Consumer test using Relative Preference Mapping ................................................................. 45 3.10.4 Data analysis of consumer acceptance test using BIB design .................................................. 46 viii University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR ....................................................................................................................................... 48 4.0 RESULTS AND DISCUSSION ........................................................................................................... 48 4.1 Focus group discussion ..................................................................................................................... 48 4.1.1 Knowledge about plant milk ...................................................................................................... 48 4.1.2 Consumption of plant and dairy milk ........................................................................................ 48 4.1.3 Marketing of plant milk ............................................................................................................. 51 4.1.4. Usage of plant milk ................................................................................................................... 52 4.1.5 Raw materials for preparation of plant milk .............................................................................. 53 4.2 Characterization of optimized products ............................................................................................ 58 4.3 Physicochemical characterization ..................................................................................................... 58 4.3.1 Proximate composition of 3-blend, commercial plant milk and dairy beverages ...................... 58 4.3.2 Color of 3-blend, commercial plant milk and dairy beverages .................................................. 60 4.3.3 pH and titratable acidity of 3-blend, commercial plant milk and dairy beverages .................... 62 4.3.4 Rheology of 3-blend, commercial plant milk and dairy beverages ............................................ 63 4.4 Sensory Characterization of optimized beverages ............................................................................ 65 4.5 Comparing sensory acceptability of optimized 3-blend beverages with commercial dairy and plant milk beverages using RPM ..................................................................................................................... 65 4.5.1 Comparison of mean liking scores of 3-blend beverages with commercial products ................ 65 4.6 Consumer acceptance test of all 19 formulation prototypes using the BIB design ........................... 70 4.6.1 Mean liking scores from consumer test ..................................................................................... 70 4.6.2 Effect of the ingredients on the liking scores for the overall liking, appearance, flavor, mouth feel, consistency and aftertaste ............................................................................................................ 73 4.6.4 Analysis of comments from the Balance Incomplete Block Design .......................................... 85 5.0 CONCLUSIONS ................................................................................................................................... 88 6.0 RECOMMENDATIONS ...................................................................................................................... 89 7.0 REFERENCES ..................................................................................................................................... 90 APPENDIX 1 ............................................................................................................................................ 100 Recruitment questionnaire used for Focus Group Discussion .............................................................. 100 APPENDIX 2 ............................................................................................................................................ 104 Moderators Guide for focus group discussion ...................................................................................... 104 APPENDIX 3 ............................................................................................................................................ 107 Recruitment Questionnaire for consumer tests (RPM and BIB) ........................................................... 107 ix University of Ghana http://ugspace.ug.edu.gh APPENDIX 4 ............................................................................................................................................ 110 Parental assent form for RPM ............................................................................................................... 110 APPENDIX 5 ............................................................................................................................................ 113 Consent form for RPM-Adult ............................................................................................................... 113 APPENDIX 6 ............................................................................................................................................ 117 Ballot Sheet for BIB .............................................................................................................................. 117 APPENDIX 7 ............................................................................................................................................ 120 Laboratory Trial of RPM ...................................................................................................................... 120 APPENDIX 8 ............................................................................................................................................ 124 Events used for RPM in social setting, pictures of RPM and ANOVA table of RPM in social and laboratory settings ................................................................................................................................. 124 x University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure 2.4 1 General processing steps for plant milk Source: Jeske et al. (2018); Makenin et al. (2016). . 13 Figure 2.5 1 Process flow chart for plant milk. Source Odoom (2018) ...................................................... 22 Figure 3.1 1 Study design ........................................................................................................................... 27 Figure 3.5 1 Process flow diagram for plant milk processing ..................................................................... 35 Figure 3.8 1 T-Map scale ............................................................................................................................ 42 Figure 4.3 1 Color of the different plant milks used in the formulations .................................................... 62 Figure 4.3 2 Effect of shear rate on apparent viscosity of products ............................................................ 65 Figure 4.5 1 Comparison of Mean liking scores (Y-Axis) between 3-blend products and commercial samples ........................................................................................................................................................ 66 Figure 4.5 2 Product map for the social setting .......................................................................................... 68 Figure 4.5 3 Product map for Laboratory setting ........................................................................................ 69 Figure 4.6 1 Cox response trace plot for overall liking ............................................................................... 73 Figure 4.6 2 Cox response trace plot for appearance .................................................................................. 75 Figure 4.6 3 Cox response trace plot for flavor .......................................................................................... 76 Figure 4.6 4 Cox response trace plot for mouth feel ................................................................................... 77 Figure 4.6 5 Cox response trace plot for consistency ................................................................................. 78 Figure 4.6 6 Cox response trace plot for aftertaste ..................................................................................... 79 Figure 4.6 7 Overlaid contour plot for formulation when peanuts milk is held at 0%................................ 80 xi University of Ghana http://ugspace.ug.edu.gh Figure 4.6 8 Overlaid contour plot for formulation when melon seeds milk is held at 0% ........................ 81 Figure 4.6 9 Overlaid contour plot when tiger nut milk is held at 0% ........................................................ 81 Figure 4.6 10 Correspondence Analysis for 3- blend formulations ............................................................ 86 Figure 7 1 Liking scores for laboratory trial of RPM ............................................................................... 122 Figure 7 2 Product map for Laboratory trial ............................................................................................. 123 Figure 8 1 Assessors using the T-map scale to evaluate plant milk in social setting RPM ...................... 125 Figure 8 2 Assessors evaluating plant milk in social event RPM ............................................................. 125 Figure 8 3 Assessors evaluating plant milk in the laboratory setting ....................................................... 126 xii University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 2 1 Classification of plant milk based on source .............................................................................. 12 Table 2 2 Proximate composition of initial plant milk ............................................................................... 23 Table 2 3 Color indices of initial plant milk ............................................................................................... 24 Table 3 1 Products tasted during focus group ............................................................................................. 30 Table 3 2 Formulations from mixtures design ............................................................................................ 31 Table 3 3 Products used for RPM ............................................................................................................... 40 Table 4 1 Proximate composition of beverages .......................................................................................... 58 Table 4 2 Color indices of beverages .......................................................................................................... 60 Table 4 3 pH and titratable acidity of beverages......................................................................................... 62 Table 4 4 Rheological parameters of beverages ......................................................................................... 63 Table 4 5 Mean liking scores for 3-blend products..................................................................................... 71 Table 4 6 Optimum formulations from overlaid contour plots ................................................................... 82 Table 4 7 Predicted liking scores at optimum regions ................................................................................ 83 Table 7 1 Products used for laboratory trial .............................................................................................. 121 Table 8 1 Events used for RPM in social setting ...................................................................................... 124 Table 8 2 ANOVA table comparing Laboratory and social setting RPM ................................................ 126 xiii University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE 1.0 INTRODUCTION Plant milk beverages are water extracts of legumes, oil seeds, cereals and nuts that resemble cow milk in appearance and consistency (Makenin et al., 2016; Swati et al., 2016). Although plant milk beverages have been in existence for many years as part of the traditional diet for various countries, there has been renewed interest in these food sources and the market for them is increasing (Jeske et al., 2018; Aidoo et al., 2010). Plant milk beverages are normally prepared and consumed at home or sold at the local market. Plants and plant materials used for plant milk include soybean, cowpea, winged bean, peanut, melon seeds (‘‘egusi or agushie’, ‘akatewa’), rice, tiger nuts, almonds, maize, rye and quinoa (Aidoo et al., 2010; Swati et al., 2016; Jeske et al., 2016). Traditional plant milk beverages include Horchata de chufa, tiger nut milk from Spain and Asian soymilk (Jeske et al., 2018). Soymilk is the most widely consumed plant milk in the world (Makenin et al., 2016). Plant milk beverages that have gained commercial significance recently include almond milk, rice milk, coconut milk, oat milk, hemp milk and quinoa milk (Jeske et al., 2018; Makenin et al., 2016). In addition to other health claims, plant milk beverages have the advantage of being lactose and cholesterol free (Mridula & Sharma, 2015). They are thus generally seen as healthier alternatives to dairy milk and have become the preferred choice for many consumers (Jeske et al., 2018). In spite of the advantages of plant milk, they lack optimal sensory attributes for consumer acceptance and consumption. A typical complaint being the beany and painty taste experienced 1 University of Ghana http://ugspace.ug.edu.gh from legume-based plant milk (Jeske et al., 2018). Various researchers (e.g. Onning et al., 1998, Palacios et al., 2009, Palacios et al., 2010, and Makenin et al., 2015) have found that consumers prefer dairy milk over plant milk especially legume-based plant milks for this reason. It has been suggested in the literature that some ways to reduce the off-flavors include processing options like roasting (Makenin et al., 2016), soaking in NaHCO3 (Diarra et al., 2005) or by the addition of natural or artificial flavors (Swati et al., 2016). Both Swati et al. (2016) and Aidoo et al. (2010) have also suggested that optimum sensory and physicochemical properties can be achieved by combining two or more legumes and oil seeds to leverage on their different physicochemical and sensory properties. In Ghana, there is an abundance of legumes, nuts and oil seeds that have potential to be transformed to plant milk. Some of these plants are underutilized and inexpensive and will benefit from being processed into new food products including milk. From a nutritional viewpoint, this will be beneficial to the consuming populace as currently animal sourced milk in Ghana is expensive and not always readily available (Otoo, 2017). Preliminary work has been done to develop plant milk beverages (which included one, two and three blend beverages) using melon seed, peanut and coconut based on the perceived health benefits of the plants, their low cost and availability in Ghana (Odoom, 2018). However, the sensory profile of the initial plant milk beverages had sensory descriptors (e.g. boiled groundnuts, spicy, groundnut soup, astringent and particulate) that suggested it will be unappealing to consumers creating the need for a re-work of those formulations. According to Pelsmaeker et al. (2015), for food product development to be more successful there should be consumer input right from the beginning. This could be in the form of qualitative methods like focus group discussions at the beginning of the process and consumer acceptability 2 University of Ghana http://ugspace.ug.edu.gh testing at the latter part to characterize the products. Food product development is time consuming and expensive, thus there is the need for a method that will save time, money and give useful information to the product developer (Ekpong et al., 2006). Relative Preference Mapping (RPM) is a consumer-based food product development tool which is based on the principle of projective mapping to place products on a two-dimensional map. With RPM, consumers decide simultaneously how much they like a product and how different the product is compared to a known reference product. The method is rapid and the researcher can quickly visualize in two dimensions, products that are liked and different to the reference product in a space called the “area of innovation” (Blay, 2012). In spite of its usefulness in food product development by identifying innovative products, it is not able to identify what factors or combination of factors lead to consumer acceptance. This can however be overcome by combining RPM with another consumer test and the data from the consumer test can be regressed to pinpoint which factors in the product are responsible for consumer acceptance of the innovative products. This research sought to use sensory methods to optimize the initial plant milk beverages prepared from melon seeds, peanuts and coconut and characterize the 3-blend beverages as a measure of the success of the optimization. 1.1 Rationale Consumer led food product development takes consumers current and future needs into consideration in the product development process and seeks to satisfy them hence results in the development of food products that meet consumer expectation and have added value. This leads to products that are accepted by target consumers. Previous work that used melon seeds, peanuts 3 University of Ghana http://ugspace.ug.edu.gh and coconut to produce plant milk beverages did not use consumer input in the product development process and thus resulted in products that had a sensory profile that suggested that it will not be accepted by consumers. Using a consumer-led approach from the beginning to the final stages in the optimization of these products will lead to products that will be acceptable to consumers. The 3-blend beverages were of interest in this work as researchers wanted to leverage on the different sensory and physicochemical properties of three raw materials. 1.2 Main Objective To use a consumer-focused approach to optimize initial plant milk beverages developed and characterize the sensory acceptability and physicochemical properties of the 3-blend beverages. 1.3 Specific Objectives i. Determine if the plant raw materials (melon seeds, peanut and coconut) used to formulate the initial plant milk beverages met consumer expectations. ii. Optimize the plant milk blends based on recommendations from target consumers iii. Assess and compare the consumer acceptability of the optimized products with commercial plant and dairy beverages using Relative Preference Mapping (RPM). iv. Determine how the different plant milks influenced consumer acceptability scores of the optimized plant milk using a consumer test with a Balanced Incomplete Block Design (BIB) v. Assess the proximate composition and physicochemical characteristics of the optimized 3- blend plant milks. 4 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO 2.0 LITERATURE REVIEW 2.1 Consumer led product development There are various reasons for product development. These include the changing needs of consumers, new technology, ingredients, market opportunities and the need to feed an ever- increasing world population with quality and sustainable foods (Linnemann et al., 2006; van kleef et al., 2005). Most new products introduced into the market fail and to improve on the chances of success there should be integration of consumer requirements into the development of products. The product developer should also understand consumer needs and translate them in an effective way. It is important to translate consumer requirement in the words of the consumers so they are not misinterpreted (Linnemann et al., 2006; Ares et al., 2011; Pelsmaeker et al., 2015). The goal of consumer led food product development is to provide foods that are acceptable to consumers. Consumers should be at the starting point of new product development. The target consumers should be identified and their needs met (Costa & Jongen, 2006). 2.1.1 Stages in product development According to van kleef et al. (2005), the main stages of food product development are the opportunity identification stage, development of the opportunity into a product, testing or optimization and launch of the product. Consumer involvement plays a very important role in the initial and final stages of food product development. In the beginning, qualitative methods could 5 University of Ghana http://ugspace.ug.edu.gh be used to understand future users. At the last stage consumers may be used to evaluate the product to see if the final product meets consumer expectations. Whilst in the intermediate stages, the design team uses the needs, wishes and desires of consumers to formulate products (Scifferstein, 2015; van kleef et al., 2005). 2.1.2 Consumer research methods in product development To attempt to use only quantitative sensory tests to find consumer response to food will not lead to the creation of a successful product. Combining qualitative and quantitative consumer research will be necessary. This is because qualitative methods help to understand consumer needs and responses. Also, approaches that take into consideration the products role in everyday life will be more beneficial to improve the validity of the product preference and liking scores (Scifferstein, 2015). Examples of qualitative research methods used in the early stages of food product development include empathic design, category appraisal, co-joint analysis, focus group, free elicitation, information acceleration, Kelly repertory grid, laddering, lead user technique and zaltoma metaphor elicitation technique (van Kleef et al., 2005). A focus group discussion consists of a homogenous group of 6-12 consumers of a particular product who meet together under the guidance of a moderator to discuss a particular topic or topics. They are used in the early stages of product development to explore new concepts and identify new opportunities. They are normally interactive discussions. The focus group is often unstructured in nature and the moderator has to probe from the answers given to get further insight into consumer needs and preferences. The responses are normally unstructured and the researcher summarizes what is said and makes inferences (van Kleef et al., 2005; Meilgaard, et al., 2006). 6 University of Ghana http://ugspace.ug.edu.gh A consumer acceptance test is a quantitative test used in the final stage of food product development. Also known as an affective test, it measures the degree to which a product is liked or disliked. The 9-point hedonic scale is the most common and popular scale to measure acceptance and preference. As the name suggests, it consists of 9 categories which range from dislike extremely to like extremely. There are various forms of the 9-point hedonic scale, some are labeled at their ends with dislike extremely and like extremely, some have only words, whilst some have only numbers (Hein et al., 2008; Meilgaard et al., 2006). 2.2 Product Optimization and Characterization The goal of product optimization is to increase the consumer acceptance of any food product being developed. Mostly, food product development has a great workload and short time frame, so there is the need to look for the best method that will save time (Ekpong et al., 2006). Product optimization involves the identification of key product attributes that are important to consumers. Optimization could also mean using sensory evaluation or other techniques to identify the best possible combination of factors that will lead to consumer acceptability of a product (Ekpong et al., 2006; Aidoo et al., 2012). In order to do this, food product developers need to understand which sensory properties drive liking or acceptance for a particular food product. There has been a school of thought that consumers are not always able to say why they like a product (Ares et al., 2010a). Because of this, preference mapping has been used in food product development to understand the sensory attributes of food that drive consumer preferences (Thompson et al., 2004). In preference mapping, a trained sensory descriptive panel is used to generate the sensory characteristics of a 7 University of Ghana http://ugspace.ug.edu.gh product and this is correlated to consumer liking ratings. This can guide food product optimization (Ares et al., 2011; Ares et al., 2010a; Thompson et al., 2004; Nunes et al., 2012). Preference mapping is detailed, reliable and reproductive (Cadena et al., 2014). One limitation of preference mapping is that consumers are only asked about their liking but how they perceive the sensory characteristics of the products is not gathered (Ares et al., 2011), it is assumed that consumers and trained assessors perceive products in the same way. However, trained panelist may describe products differently or consider attributes that are irrelevant to consumers (Ares et al., 2010, Ares et al., 2011). Another limitation is that, the use of trained assessors to create a sensory profile of the products is very time consuming and expensive, making the use of preference mapping in food product development laborious and expensive (Ares et al., 2011; Bruzzone et al., 2015; Cadena et al., 2014). As reviewed earlier, most researchers emphasize the implementation of a consumer-oriented approach as the best way to develop successful new products. This involves gaining a deep understanding of the ‘voice of the consumer’ and incorporating consumer opinions in every stage of the food product development process (Bruzzone et al., 2015; Linnemann et al., 2006; Ares et al., 2011; Pelsmaeker et al., 2015; Costa & Jongen., 2006). Recently, there has been an interest in letting consumers give a description of the sensory characteristics of products. Researchers have advocated for the use of consumer sensory characterization in product development as an alternative to the classical sensory characterization using Quantitative Descriptive Analysis (QDA) (Dooley et al., 2010; Ares et al., 2011). This is to overcome the limitations of preference mapping. Another reason is that the best way to understand consumer preference is through consumers themselves (Ares et al., 2011). Consumer based sensory characterization methods help to bring consumer perception of products into the 8 University of Ghana http://ugspace.ug.edu.gh food product development process. Because they do not require training and maintaining a sensory panel, they are less time consuming and less expensive and are thus known as rapid methods (Antunez et al., 2017). Consumer characterization methods such as intensity scales, Just about right (JAR) scales, projective mapping, flash profiling, check-all-that-applies (CATA), polarized sensory positioning (PSP) have been designed to provide product sensory characterization in a short time frame (Ares et al., 2011; Cadena et al., 2014; Bruzzone et al., 2015). Consumer characterization using CATA involves giving consumers products and providing a set of sensory descriptors. They are instructed to check all the descriptors that describe the product. No intensities are given to the attributes and descriptors are not constrained to only sensory attributes but usage and concept fit (Dooley et al., 2010; Ares et al., 2015; Antunez et al., 2017). The advantages of this method are that it provides information about which attributes are detectable by consumers and how they relate to their overall liking and acceptance (Dooley et al., 2010). Because it gets rid of scaling, it is easy, natural and intuitive to consumers (Bruzzone et al., 2015, Ares et al., 2011). According to Cadena et al. (2014), CATA is simple, valid, reproducible and in a study to compare CATA, projective mapping, polarized sensory positioning (PSP) to the traditional QDA in the development of functional yogurt, CATA gave the most similar results to QDA. A limitation of CATA is that it produces frequencies rather than rating or intensity, because of this it has less power than quantitative data and requires a lot of consumers. Another issue with CATA is how to optimize the list of descriptors in order to produce more efficient descriptors (Valentin et al., 2012) In projective mapping, consumers are provided with all products at the same time and asked to place them on an A3 sheet based on their similarities and differences using their own criteria. 9 University of Ghana http://ugspace.ug.edu.gh Products positioned close together are considered similar whilst those far apart are considered different. It can be done using a trained or untrained panel (Ares et al., 2015; Antunez et al., 2017, Ares et al., 2010b). Projective mapping could be used to produce perceptual maps similar to those obtained with descriptive analysis for the obvious aspects of products and results are reproducible at group level (Valentin et al., 2012). It is simple and useful to evaluate perception of food products, but the differences between products are difficult to explain (Ares et al., 2010b). Limitations include difficulty in constraining assessors to use two dimensions to discriminate between products, the need to present products at the same time, as in the case of unstable products like ice cream and difficulty in interpreting descriptions provided by assessors (Valentin et al., 2012; Blay, 2012) With polarized sensory positioning (PSP), consumers are given three references (poles), which they evaluate first and then are asked to evaluate the products based on differences between them and each of the poles using an unstructured scale anchored with totally different and exactly the same using their own criteria (Ares et al., 2015; Antunez et al., 2017). PSP is easy to perform as assessors are only required to evaluate global similarities and differences between samples. It also allows the compilation of data across sessions. In spite of its advantages, its limitations are that, descriptions of products can only be obtained indirectly by deduction from the sensory characteristics of the most similar reference, it can only be used to aggregate data only if the references are stable and it requires a good knowledge of the product space in order to choose optimal references (Valentin et al., 2012) Relative Preference Mapping (RPM) was developed based on the principle of projective mapping. One difference between this method and projective mapping is the use of a reference sample (this makes it similar to PSP) and another is that researchers define the axes rather than 10 University of Ghana http://ugspace.ug.edu.gh letting consumers choose their own attributes to differentiate products on the map. The two axes used in RPM are the ‘liking’ and ‘difference’ axes. Consumers place products on the two- dimensional map by simultaneously evaluating how different the product is and how much they like the product compared to the reference. One major advantage of this characterization method is that it enables the product developer to rapidly visualize innovative products in an “area of innovation” between the liking and difference axes. It is also easy to perform by assessors because products are compared to a reference sample and it is easier to make relative judgments than absolute judgments. But a disadvantage with this method is that, like PSP the product developer is only able to deduce the sensory characteristics of the samples from the reference and is not able to pinpoint what sensory factors are responsible for preference (Blay, 2012). RPM was designed to be used in a social setting. The evaluation is designed to be like a game, taking into consideration the ambience of the test location. The results from RPM in a social setting and laboratory setting have not been compared. Given its use as a product development tool in this research, it is important to test the results of RPM in the lab and in social setting to determine if comparable results can be obtained (Blay, 2012). According to Meiselman (2013), consumer preference testing in the normal consumption environment of consumers is more reflective of consumer preference. 2.3 Emerging consumer trends leading to plant milk product development Consumers have now become more interested in foods that contribute to their health and well- being (Saalia et al., 2013; Swati et al., 2016). Issues such as lactose intolerance, cow milk allergy, high cholesterol, vegetarianism have made the introduction of plant milk necessary 11 University of Ghana http://ugspace.ug.edu.gh (Swati et al., 2016; Aidoo et al., 2012; Makenin et al., 2016; Jeske et al., 2016). Ghana does not produce enough animal-sourced milk but imports for food manufacture and consumption. The availability of plant milk will lead to an affordable option for Ghanaians (Saalia et al., 2013). Soymilk is the most popular plant milk and has received a lot of research interest and has dominated the market for plant milk. But attention has been drawn to using other plants for plant milk. It is expected that the market for plant milk in Ghana and the world at large will increase in the future (Saalia et al., 2013; Swati et al., 2016). 2.4 Plant Milk According to Makenin et al. (2016), plant milk beverages are extracts of plant material including legumes, oil seeds and pseudo-legumes which resemble milk in appearance. They are also described by the same authors as colloidal suspensions or emulsions comprising of disintegrated or dissolved plant material. Swati et al. (2016), describe plant milk as the fluid that results when plants are extracted in water and homogenized. This fluid resembles cow’s milk in appearance and consistency. Table 2.1 shows the classification of plant milk based on its source. Table 2 1 Classification of plant milk based on source Classification Examples Cereal based Oat milk, Corn milk, Spelt milk, Rice milk Legume based Soy milk, Peanut milk, Lupin milk, Cowpea milk Seed based Sesame milk, Flax milk, Hemp milk, Sunflower milk Nut based Almond milk, Coconut milk, Pistachio milk, Walnut milk Pseudo-cereal based Quinoa milk, Amaranth milk, Teff milk Source: Swati et al. (2016). Plant based Beverages: An emerging Segment. 12 University of Ghana http://ugspace.ug.edu.gh 2.4.1 Preparation of Plant milk Generally, plant milks are prepared by sorting out the raw material to remove defective and contaminated material, washing the raw material with water and then blending the plant material with the required ratio of water to get a slurry and then straining to get the milk. The manufacturer may add sugar, stabilizer, flavorings and oil. After this, homogenization and pasteurization follow to improve the stability and shelf life respectively (Makenin et al., 2016; Jeske et al., 2018). Figure 2.4.1 shows the basic processing steps in plant milk preparation. Grains/pulses/nuts/seeds Soaking Dry Milling Wet Milling Extraction of flour Separation of solids (Filtration or decanting) Product Formulation (Standardization, addition of additives) Homogenization Pasteurization/UHT Packaging Figureure 2.1 1 Plant milk Figure 2.4 1 General processing steps foFrig uprleaunrte m 2.i1lk 2 Source: Jeske et al. (2018); Makenin et al. (2016). 13 University of Ghana http://ugspace.ug.edu.gh 2.4.2 Advantages of consuming plant milk Plant milk beverages have many health benefits. Oats milk for instance, lowers cholesterol (Onning et al., 1998) and has cancer fighting properties (Swati et al., 2016). Β-glycan in oats is responsible for lowering the blood sugar level (Swati et al., 2016). Oat milk has antioxidants and polyphenols (Swati et al., 2016). Soy milk is the most popular plant milk (Makenin et al., 2016). It is a good source of monosaturated and polysaturated fatty acids (Swati et al., 2016). Soymilk protects against cancer, heart disease and osteoporosis because of its isoflavone content (Omoni & Aluko, 2005). It also lowers cholesterol (Swati et al., 2016: Makenin et al., 2016). Almonds are a good source of vitamin E (Swati et al., 2016; Dhakal et al., 2014). Almond milk also has prebiotic properties (Swati et al, 2016) and may reduce the total cholesterol level in the blood (Swati et al., 2016; Dhakal et al., 2014). Apart from this, it helps with free bowel movement, has antioxidant properties and improves immune function (Swati et al., 2016). Coconut milk can be easily digested (Aboulfazli et al., 2014). It helps in healing mouth ulcers and contains lauric acid (a saturated fatty acid) which promotes brain development (Ganguly, 2013; Belewu & Belewu, 2007). In addition to lauric acid, it also has high oleic acid content, these fatty acids help in the prevention of arteriosclerosis and other heart related ailments (Aboulfazli et al., 2014). The vitamin E content in coconut milk makes it a good source of antioxidants that helps fight against aging, boost immune function and strengthen blood vessels by making them more flexible (Swati et al., 2016). Coconut milk has a cooling effect on the body, helps with digestion and nourishes the skin (Swati et al, 2016). Tuso et al. (2013), advocate for a shift towards a plant-based diet, according to them, a plant-based diet maximizes the consumption of nutrient dense plant foods whilst reducing the consumption of processed foods and animal foods. According to them, the 14 University of Ghana http://ugspace.ug.edu.gh benefits of a plant-based diet include the reduction of obesity, heart disease, high blood pressure, diabetes and cancer. 2.4.3 Disadvantages of Plant milk Plant milk beverages are colloids with large sized dispersed phase particles which can sediment to form unstable suspensions. The large particles also result in the sandy, gritty and chalky mouth feel often experienced with plant milk (Swati et al., 2016). The stability can be improved by reducing the particle size and using hydrocolloids, stabilizers and emulsifiers. Good homogenization also improves stability by reducing particle size, disrupting aggregates and lipid droplets to prevent sedimentation and separation in the plant milk beverage. This can be achieved by the use of the colloid mill (Makenin et al., 2016; Swati et al., 2016). Heating and pasteurization also lead to sedimentation, gelling or formation of aggregates. Normally, when plant milk is heated, the non-polar amino acids come into contact with water and increases their hydrophobicity, this increases the interactions between proteins and leads to sedimentation or gelling which leads to increase in viscosity (Makenin et al., 2016). This is reduced by using a protease to break down the proteins into peptides or amino acids to prevent their aggregation and the resultant increase in viscosity. One drawback however, is the production of bitter peptides which makes the hydrolysates taste bitter (Humiski & Aluko, 2007). Another phenomenon that leads to increase in viscosity is the gelatinization of starch; this is solved by enzyme hydrolysis of the starch using alpha amylase (Swati et al., 2016). Plant milk beverages are often low in protein and calcium (Makenin et al., 2016; Jeske et al., 2016). The nutritional content of plant milk depends on the raw material used, processing, 15 University of Ghana http://ugspace.ug.edu.gh fortification and addition of sweeteners and oil. Water soluble vitamins can be lost if the raw material is soaked or blanched before production. Minerals can also be lost during the decanting step of processing. Heat treatment can also destroy some heat sensitive vitamins and denature proteins depending on the temperature and exposure time (Makenin et al., 2016). Some plants often contain some antinutritive properties such as inositol phosphates (phytates) and trypsin inhibitors. Inositol phosphates bind to calcium, zinc, magnesium and iron and reduce their bioavailability. Trypsin inhibitors lower the digestibility of protein but they can be inactivated by heat treatment (Makenin et al., 2016; Swati et al., 2016). Another disadvantage of plant milk is the low consumer acceptability of plant milk. Many consumers have described plant milk beverages as having a chalky mouth feel. Legume based plant milk have been described as being “painty” and “beany”. For instance, in a study conducted by Onning et al. (1998), acceptability scores of oats, cow and soy milks were compared; oat milk was preferred over cow’s milk, but soymilk had the lowest acceptability score. Palacios et al. (2009) also compared the acceptability of lactose free cow’s milk and soymilk and found lactose free cow’s milk was preferred over soymilk. Similar results were found when the same researchers tested those products with American school children the following year (Palacios et al., 2010). In these three studies, it can be seen that soymilk (produced from a legume) had low acceptance and in the first study was less accepted compared to oat milk (produced from a cereal). Aboulfazli et al. (2014) studied the acceptability of ice cream in which the skimmed milk powder had been partially or completely replaced by either soymilk or coconut milk, ice creams containing skimmed milk and a lower percentage of vegetable milk were the most accepted. With ice creams containing only soymilk and coconut milk, those ice creams with a higher percentage of coconut milk had higher acceptance scores for 16 University of Ghana http://ugspace.ug.edu.gh flavor. Processing methods like roasting (Makenin et al., 2016), blanching (Makenin et al., 2016) and soaking in alkaline (Diarra et al., 2005) can be used to reduce beany flavor in legumes. Addition of sugar and other flavors also increase the acceptability of plant milk (Swati el al., 2016). The taste and nutritional deficiencies of oil seeds and legumes can be overcome by combining two or more oil seeds and legumes (Swati et al., 2016; Aidoo et al., 2010). Aidoo et al. (2010), combined peanuts and cowpea to improve the acceptability of a vegetable milk powder used to replace skimmed milk in chocolate manufacture. Adding the peanuts to cowpeas reduced the beany flavor of the cowpeas. 2.4.2 The use of local plants in the commercial processing of plant milk in Ghana To be able to produce plant milk on a commercial scale in Ghana, raw materials must be sourced locally. This will lead to a reliable and sustainable supply of raw materials. This is because local raw materials are available, can be grown here and are often not expensive. Examples of potential raw materials include coconuts, melon seeds, tiger nuts, peanuts, cowpea, almonds, and rice. Coconut (Cocos Nucifera L) referred to as the tree of life has various uses and is used by millions of people around the world (Pham, 2016). Coconut can be processed into oil and milk, its water is also a refreshing beverage (Ganguly, 2013). Raghavendra and Raghavarao (2010) describe coconut milk as an oil in water emulsion which contains 54% water, 35% fat and 11% milk solids non-fat. The milk solids non-fat is made up of 2-4% protein (Tansakal & Chaisawang, 2006). Coconut milk is produced from grating the copra of the coconut (Cocos nucifera). It is a good substitute for cow’s milk because it is easy to prepare, is easily digested and is nutritious. Coconut milk is rich in minerals such as phosphorus, iron, magnesium, zinc, 17 University of Ghana http://ugspace.ug.edu.gh calcium and potassium and vitamins C, E and many B vitamins (Swati et al., 2016; Aboufazli et al., 2016). It is also rich in glutamic, aspartic acid, arginine, lipids and antioxidants. The fatty acids, lauric and oleic acid present in coconut milk help in the prevention of atherosclerosis (Aboulfazli et al., 2016) Melon seeds (Colocynthis Citrullus L) locally known as ‘agushie’ or ‘egusi is the biological ancestor of the water melon. It is an oil seed consumed mainly in West Africa (Bankole et al., 2005). It is now found all over the world but particularly in West Africa, where it is part of daily meals (Ojieh et al., 2008; Lawal et al., 2012; Oluba et al., 2008). The seeds of this plant have been shown to decrease glucose levels in diabetic patients. Melon seeds are used for the treatment of constipation, edema, infections, and rheumatism and it serves as a natural purgative (Ezuruike & Prieto, 2014). Melon seeds are rich in essential fatty acids and minerals (Adjepong et al., 2017). According to Adjepong et al. (2017), they should be incorporated into the Ghanaian diet to help in the prevention of nutritional deficiencies of essential fatty acids and minerals such as iron, zinc and manganese which could lead to growth stunting and cognitive impairment as the Ghanaian diet is not diverse. Melon seeds are used as thickeners and flavoring in soups (Abiodun and Adeleke, 2010) and serve as emulsifiers and fat binders in soups (Akubor et al., 2003). They are also used to thicken stews (Oluba et al., 2008). Melon seeds oil is edible (Bankole et al., 2005). Another use for melon seeds is for milk as an alternative for soybeans (Akubor, 1998). Melon seeds milk prepared had 3.5% protein, 4.0 % fat and 2.5% carbohydrate (Akubor et al., 2003). Peanut (Arachis hypogaea) also known as groundnut, is a pea that belongs to the Fabaceae family which is legumes/beans. Because of its high oil content, it is also considered as oilseed (Arya et al., 2016; Jain et al., 2013). Peanuts are known to help with weight loss when high 18 University of Ghana http://ugspace.ug.edu.gh energy elements in the diet are replaced with them. They are able to do this because they induce satiety and lead to a lower energy intake in subsequent meals. This is because nuts are rich in fiber and protein. Studies have shown that the consumption of nuts lead to high resting energy expenditure (Martinez-González & Bes-Rastrollo, 2011). Peanut milk is rich in proteins, minerals, essential fatty acids such as oleic and linoleic acid (Isanga & Zhang, 2009). It has low levels of calcium, copper and zinc but high levels of potassium, magnesium, sodium and phosphorus (Kane et al., 2010). Tiger nuts (Cyperus esculentus L) is an edible perennial grass like plant that belongs to the old world, it is also known as earth almond, ‘chufa’ (in Spanish), earth nut, yellow nut, sedge, groundnut, rush nut, edible galingale and nut grass (Sánchez-Zapata et al., 2012; Ayeh-Kumi et al., 2014; Ogunlade et al., 2015; Sebastià, et al., 2012). The tiger nut plant produces sweet, nut- like tubers which are energy dense (Codina-Torrella et al., 2015). Tiger nuts are rich in sugar, protein, fatty oil, fiber and in minerals such as sodium, calcium, potassium, phosphorus, magnesium, zinc, traces of copper, vitamins E and C (Aguilar et al., 2015; Sebastià, et al., 2012; Codina-Torrella et al., 2015; Hamdy et al., 2017). Their fatty acid profile is similar to olive and hazel nuts (Lasekan & Abdul Karim, 2012; Albors et al., 2016; Codina-Torrella et al., 2015). Tiger nuts have anti- inflammatory properties (Hamdy et al., 2017). It is an underutilized tuber in many parts of the world but is used to prepare a cold beverage called Horchata in Spain (Aguilar et al., 2015). Tiger nuts are suitable for people with diabetes, digestive disorders and cardiovascular diseases (Codina-Torrella et al., 2015). Tiger nuts are a popular snack in Ghana. They are either eaten raw, roasted or the milk from the nuts is boiled into porridge. They can also be ground to make beverages (Owusu Yeboah, et al., 2012; Ezeh et al., 2014). They can be processed into flour for baking (Albors et al., 2016). Tiger 19 University of Ghana http://ugspace.ug.edu.gh nut milk is the aqueous extract obtained from the tiger nut tubers (Kizzie-Hayford et al., 2016). Tiger nut milk is made up of 0.91% protein, 0.25% ash, 3.09% fat and 1.03% dietary fibre (Sánchez-Zapata et al., 2012). 2.5 Preliminary work on the initial plant milk beverages Odoom (2018) used melon seeds, peanuts, and coconut to formulate plant milk beverages after which physicochemical and sensory properties were characterized. The raw materials were chosen because of their perceived health benefits, availability and affordability. Other reasons included adding value to them and the prevention of post-harvest losses in Ghana through their utilization. Using peanuts and melon seeds was to reduce the aflatoxin content of these plant materials and enhance usability. The two raw materials are also complementary to each other in their essential amino acid composition. The milk formulation was achieved using a three component simplex mixtures design in Minitab with the three raw materials as the components. This resulted in 10 formulations, which were one, two and three blends. In processing the beverages, plant milk from each of the raw materials was produced separately and blended together according to the ratios generated from the mixtures design. For peanut milk preparation, peanuts were sorted to remove contaminated and defective seeds, blanched in boiling water for 3 minutes, rinsed in cold water and then dehulled. The dehulled peanuts were then steeped in 2% NaHCO3 for 24 hours and washed with clean water. They were blended with water in the ratio of 1:4 (w/v) in a commercial blender and passed through colloid mill to obtain milk. Sugar, buffer and stabilizer were added and the milk was heated to 70oC to dissolve the additives. Melon seeds milk was processed in this way; the 20 University of Ghana http://ugspace.ug.edu.gh melon seeds were sorted to remove discolored and contaminated seeds. Melon seeds were roasted at 1200C for 10 minutes in an electric oven. The roasted seeds were blended with water in a ratio of 1:5 (w/v) in a commercial blender and passed through colloid mill to obtain melon seeds milk. Sugar, buffer and stabilizer were added and the milk was heated to 70oC to dissolve the additives. Lastly, for coconut milk production, the coconuts were shelled and pared using the method described by Seow et al. (1997). The copra was washed in clean water and chopped into pieces. The chopped coconut was blended with water using a commercial blender (a ratio of 1:3 (w/v). The slurry was then filtered using a cheese cloth which was manually twisted to obtain the coconut milk. Sugar, buffer and stabilizer were added and the milk was heated to 70oC to dissolve the additives. The three pretreated milk samples were then mixed according to the ratios obtained using the three-component simplex mixture design. Each of the ten samples was pasteurized at 800C for 20mins and hot filled into chlorine (50ppm) sterilized 300ml thick plastic bottles with caps. The samples were allowed to cool to room temperature and then stored in a refrigerator set at 50C. Figure 2.5.1 shows the process flow diagram for the processing of the plant milk beverages. Melon seeds milk and formulations containing melon seeds became viscous during the pasteurization of the beverages. Melon seed extracts swell up when heated, this is due to the denaturation of its proteins when exposed to heat and above the denaturation temperature gels are formed and it remains in this state even after cooling (Uruakpa & Aluko 2004). 21 University of Ghana http://ugspace.ug.edu.gh Peanuts Melon seeds Coconuts Sorting Shelling Sorting and blanching Paring 0 Dehulling & sorting Roasting @ 120 C (10mins) Blending Steeping in 2% NaHCO3 (24hrs) Blending and milling Sieving Blending and milling Milk Milk Milk Buffer, Buffer, Buffer, stabilizer stabilizer stabi lizer and sugar and sugar and sugar 0 Heat @ 70 C Heat @ 700C Heat @ 700C Peanut Milk Melon seed milk Coconut milk Blend according to ratios from mixtures design H o m o g e n i z e P a s t e u r i z e @ Plant milk beverages 800C (20mins) Figure 2.5 1 Process flow chart for plant milk. Source Odoom (2018) 22 University of Ghana http://ugspace.ug.edu.gh The physicochemical characteristics including proximate composition, emulsion stability, color, titratable acidity and rheological characteristics were determined. The shelf life of the beverages was also measured. Tables 2.2 and 2.3 show the proximate composition and color indices for the plant milk respectively Table 2 2 Proximate composition of initial plant milk SAMPLE TOTAL SOLIDS PROTEINS CHO FAT ASH (%) (%) (%) (%) (%) 1(P100) 20.41 ± 0.14 a 6.37 ± 0.78b 6.10 ± 0.08a 9.01 ± 0.13c 0.42 ±0.01ef 2(P b75M150C75) 15.00 ± 0.10 6.59 ± 0.03 a 2.13 ± 0.15ef 6.95 ± 0.09f 0.39 ±0.01gh 3(P100M100C100) 16.72 ± 0.06 ab 4.95 ± 0.08c 3.90 ± 0.00bc 8.51 ± 0.14d 0.45 ± 0.00de 4(P C ) 17.89 ± 0.09ab 4.69 ± 0.04f 3.48 ± 1.11bcd150 150 10.89 ± 0.13 a 0.38 ± 0.00gh 5(M ) 13.38 ± 0.05 b 5.83 ± 0.04c300 1.18 ± 0.05 f 7.04 ± 0.09ef 0.40 ± 0.02cd 6 (C300) 13.26 ± 0.08 b 4.94 ± 0.04e 0.99 ± 0.08f 8.10 ± 0.12d 0.36 ± 0.00h 7 (M C ) 14.73 ± 0.06b150 150 5.95 ± 0.07 c 1.85 ± 0.07ef 7.48 ± 0.11c 0.47 ± 0.00cd 8(P M C ) 16.13 ± 4.05ab 5.26 ± 0.04d 4.87 ± 0.38ab 9.92 ±0.22b 0.50 ± 0.01b150 75 75 9(P150M150) 16.72 ± 0.21 ab 6.36 ± 0.01b 2.87 ±0.08cde 8.13 ±0.07d 0.55 ± 0.01a 10 (P ab g cde b bc 75M75C150) 16.83 ± 0.05 4.34 ± 0.03 2.98 ± 0.07 10.06 ±0.08 0.49 ± 0.01 23 University of Ghana http://ugspace.ug.edu.gh Table 2 3 Color indices of initial plant milk Sample L a b 1(P100) 81.51± 0.16 ab -2.26 ± 0.10cd 6.10 ± 0.02f 2(P75M150C100) 79.48± 0.16 bc -2.00 ± 0.06bcd 8.23 ± 0.13d 3(P M C ) 79.69 ± 2.92d 0.12 ± 0.13a100 100 100 7.48 ± 0.52 e 4(P150C150) 83.86 ±0.21 a -2.76 ± 0.04d 4.73 ± 0.06 g 5(M ) 78.29 ± 0.16cd300 -1.16 ± 0.11 bc 9.44 ± 0.13b 6 (C300) 84.00 ± 0.16 a -1.78 ± 1.22bcd 0.85 ± 0.14h 7 (M150C150) 78.26 ± 0.24 cd -1.31 ± 0.16bc 10.02 ± 0.06a 8(P150M75C75) 79.28 ± 0.05 bcd -1.51 ± 0.06bc 8.55 ± 0.08cd 9(P150M150) 79.02 ± 0.30 bcd -1.03 ± 0.20ab 9.08 ± 0.23bc 10(P75M75C150) 79.32 ± 0.09 bcd -1.72 ±0.12bcd 8.39 ± 0.05d The products were shelf stable after a month of storage at 5OC. The sensory characteristics of the beverages were also determined using Quantitative Descriptive analysis (QDA). During the QDA, trained panelists came up with descriptors to describe the attributes of the beverages. The descriptors used to describe the appearance of the samples included white, cream, particulate, viscous and bubbly. Whilst aroma descriptors included groundnut soup, boiled ground nuts, coconut milk and spicy. Flavor descriptors included boiled groundnuts, coconut milk and spicy. Texture descriptors included smooth and viscous, whilst aftertaste descriptors included astringent, boiled groundnut, coconut milk, sweet and umami. These descriptors suggested that the beverages will be unappealing to consumers. Odoom (2018) made the following recommendations; 24 University of Ghana http://ugspace.ug.edu.gh • the process for processing the plant milk beverages be optimized, • the consumer acceptability of the beverages be tested, • a solution to be found for the gelling of melon seeds milk when heated, • determination of the amino acid profile of the preferred milk after consumer testing • bioavailability of nutrients in the beverages should be tested • the performance of the beverages as a creamer for coffee should also be tested 25 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE 3.0 METHODOLOGY 3.1 Study design This research followed a consumer led product development approach. Figure 3.1.1 shows the study design followed. The study was in two main parts; optimization and characterization. Consumer insight was used right from the beginning of the research. Focus group discussions were used to confirm the selection of the three ingredients used in the initial plant milk beverages and to understand which sensory properties of plant milk are desirable for consumers. Insight gained from the focus group discussions was used to optimize the plant milk beverages. Optimization involved 3 key elements; ingredient change (tiger nuts was added to melon seeds, peanuts and coconuts), process change and new process development (peanut was roasted instead of blanched and a process for tiger nut milk production was developed). Lastly, mixtures reformulation (minimum amount of coconut milk was kept constant at 25%, and tiger nuts was added to the three other components, making it a four component mixtures design). After optimization, the beverages were characterized by sensory methods; consumer tests using RPM (to ascertain consumer acceptance and compare the optimized products to commercial plant and dairy milk) and consumer test using BIB (to assess which of the ingredients was responsible for consumer acceptance of the beverages). The physicochemical characteristics including proximate composition, pH, titratable acidity, color and viscosity were also determined. 26 University of Ghana http://ugspace.ug.edu.gh Initial plant milk recipe Focus Group Discussion Product Optimization Ingredient Process change Reformulation Change and process development 19 –product runs Product Characterization Sensory- Physicochemical Consumer tests analysis of 3- blend samples B I B ( a l l 1 9 r u n s ) R P M u s i n g 3 b l e n d Proximate composition formulations+ Color Commercial plant milk beverage + dairy milk pH/Titratable acidity beverage Viscosity Figure 3.1 1 Study design 27 University of Ghana http://ugspace.ug.edu.gh 3.2 Methods 3.3. Focus group discussions The purpose of using the focus group discussions was to confirm the selection of the three ingredients (melon seeds, peanuts and coconut) used for the initial plant milk beverages and to understand which sensory properties of plant milk are desirable to consumers 3.3.1 Participants Participants were recruited from the University of Ghana and the University of Professional Studies campuses using one-on-one interviews (Recruitment questionnaire is presented in Appendix 1). To qualify to participate in the focus group discussions, potential participants had to be 18 years of age and above, consumers of plant milk who had consumed a plant milk beverage at least a month before the focus group. They could also be consumers of dairy milk (as the plant milk beverages were to be dairy milk-like and the target market was not restricted to only plant milk consumers). Potential participants were also to be free of any allergies and intolerances of the ingredients used in the plant milk beverages (soybeans, melon seeds, peanuts, coconut, and almonds) and dairy milk. They had to be fluent in English, the language the focus group was conducted in. Lastly, they had to be willing to come to the Sensory Evaluation Laboratory of the Nutrition and Food Science Department of the University of Ghana for the 90- minute focus group discussion. Demographic information was collected from participants after they were selected. They were called two days to the focus groups to remind them of the discussions. 28 University of Ghana http://ugspace.ug.edu.gh 3.3.2 Organization of the focus group Three focus group discussions were organized with a total of 17 participants. Participants were aged between 18-49 years and the three groups were made up of: Females aged 18 to 30 years, males aged from 18 to 30 years and a mixed group of males and females aged between 31 to 49 years. Participants signed a consent form before the focus group discussions began. The focus groups were audio recorded and there was a note-taker. Each focus group lasted for 90 minutes. The moderator used a semi-structured questionnaire and probing was used for the interviewing process where follow up questions were based on responses given. The sessions were in two parts, the first part was a discussion and the second part was a tasting session where participants tasted commercial plant milk as well as laboratory prepared melon seed milk, coconut milk and peanut milk. For the first part of the focus group, the participants explored the following themes: • Plant milk consumed • Knowledge about plant milk • Where plant milks consumed are bought from • Frequency of plant milk consumption • Reasons for plant milk consumption • Plant milk as a creamer • Market for plant milk • Use of peanuts and melon seeds for plant milk • Other plants for plant milk The following are some of the themes for the tasting session: 29 University of Ghana http://ugspace.ug.edu.gh • Initial impression about the product • Likes and dislikes about the product • Use of product as a creamer • What can be done to optimize the product? See Appendix 2 for the full moderator’s guide. 3.3.3 Products used for focus group The products tasted during the focus group are shown in Table 3.1 Table 3 1 Products tasted during focus group Food product Code Brand name Description Vitamilk Energy (Commercial F1 Soymilk sample) Almond Breeze (Commercial F2 Almond Milk sample) F3 Laboratory prepared coconut milk Coconut milk F4 Vitamilk-Regular Soymilk F5 Laboratory prepared peanut milk Peanut milk Laboratory prepared melon seed F6 Melon seed milk milk F7 Soy Milk (UHT) Soymilk F8 Lactasoy Soymilk The melon seeds, peanuts and coconut milk were freshly prepared by traditional methods using a kitchen blender and strainer. The products were brought to the participants both blind and with information. 30 University of Ghana http://ugspace.ug.edu.gh 3.4 Plant milk optimization The optimization work of this research involved 3 key elements; ingredient change, process change and new process development and finally reformulation (redesign of formulation). These changes were guided by insight from the focus group discussions. The 3 elements are outlined as follows: • Ingredient change: Tiger nut was added as a new ingredient • Process change: A new process was developed for the processing of tiger nuts and procedure for processing peanuts milk was changed (peanuts were roasted instead of blanched) • Reformulation (Redesign of formulation): The minimum quantity of coconut milk in the formulations was kept constant at 25% and tiger nuts was added to the other three components and a four component mixtures design from Minitab 17 was used for the reformulation. This resulted in 19 runs as shown in Table 3.2; Table 3 2 Formulations from mixtures design Percentage of each component (%) Formulation Product Coconut milk Melon seeds milk Peanuts milk Tiger nuts milk code 1 A 62.5 0 37.5 0 2 B 62.5 37.5 0 0 3 C 25 37.5 37.5 0 4 D 50 25 0 25 5 E 34.4 9.4 9.4 46.9 6 F 43.8 18.8 18.8 18.8 31 University of Ghana http://ugspace.ug.edu.gh 7 G 34.4 9.4 46.9 9.4 8 H 25 0 75 0 9 I 50 0 25 25 10 J 34.4 46.9 9.4 9.4 11 K 25 25 25 25 12 L 25 75 0 0 13 M 62.5 0 0 37.5 14 N 50 25 25 0 15 O 71.9 9.4 9.4 9.4 16 P 25 37.5 0 37.5 17 Q 25 0 37.5 37.5 18 R 100 0 0 0 19 S 25 0 0 75 3.4.1 Raw materials used for plant milk processing Melon seeds (Colocynthis Citrullus L) and coconuts (Cocos Nucifera L) were sourced from Madina market in Accra. Peanuts (Arachis hypogaea) were sourced from the Adenta Market, whilst tiger nuts (Cyperus esculentus L) were bought from the Kaneshie Market. The yellow colored tiger nuts were used. The raw materials were stored in the cold room of the Department of Nutrition and Food Science, University of Ghana, Legon at a temperature of 4oC. Buffers (K2HPO4 and KH2PO4) and chlorine tablets were bought from Lab Aid in Accra. Xanthan gum was bought from Micrite and Bromelain tablets from Sources Naturals in the United States of America. 32 University of Ghana http://ugspace.ug.edu.gh 3.4.2 Plant milk processing All raw materials were processed into milk and the different milks were mixed together. The processing of coconut milk followed the procedure described by Odoom (2018) with modification. During melon seed processing in the preliminary work by Odoom (2018), melon seed milk thickened during pasteurizing and led to an increase in viscosity in melon seed milk and formulations containing melon seed milk. In order to prevent this, processing of melon seeds milk was modified to include the addition of Bromelain (an enzyme) to hydrolyze the proteins in the melon seeds milk and prevent an increase in viscosity with heating. For this work, melon seeds milk was produced in this way; melon seeds were manually sorted to remove contaminated and defective seeds. They were washed with water and roasted for 20 minutes at 120oC in an electric convectional fan oven (HC 62062, Kaiser, Berlin, Germany) The melon seeds were blended with water in the ratio of 1:5 in a commercial soymilk mixer (PB103, China) for 2 minutes. The slurry was filtered using a plastic kitchen mesh strainer. 0.005 % (w/v) of Bromelain enzyme (Sources Naturals Bromelain, 2000 GDU/Gram) was added to the plant milk extract at pH of 6.6 at 28oC for 1 hour. Weighed quantities of buffer, cane sugar and stabilizer were added to the milk, heated to 70oC and was passed through a colloid mill (Premier 84, Premier Colloid Mills Limited, Walton on Thames, Surrey, UK). Peanuts were roasted instead of blanched during peanut processing. As a result, the peanuts milk was produced as follows: peanuts were manually sorted to remove contaminated and defective seeds, they were washed with water and roasted for 40 minutes at 120oC in the oven and dehulled after roasting. The dehulled peanuts were blended with water in the ratio of 1:4 in the soymilk mixer for 2 minutes. The slurry was filtered using a plastic kitchen mesh strainer. 33 University of Ghana http://ugspace.ug.edu.gh Weighed quantities of buffer, cane sugar and stabilizer were added to the milk, heated to 70oC and was passed through the colloid mill. A new process had to be developed for tiger nut milk processing which is as follows; tiger nuts were manually sorted to remove contaminated and defective nuts. They were washed with water and roasted for 15 minutes at 120oC in the oven. The tiger nuts were blended with water in the ratio of 1:4 in the soymilk mixer for 2 minutes. The slurry was filtered using a plastic kitchen mesh strainer. Weighed quantities of buffer, cane sugar and stabilizer were added to the milk, heated to 70oC and was passed through the colloid mill. To produce the various formulations, the four pretreated plant milk samples were mixed according to the ratios obtained using the four component simplex mixtures design as shown in Table 3.2. The samples were pasteurized at 80oC for 20 minutes. The pasteurized beverages were hot filled into 500ml chlorine sterilized plastic bottles with cups and closed. They were rapid chilled to 5 oC and stored at 4 oC. Figure 3.5.1 shows the process flow diagram for the processing of the plant milk beverages. 34 University of Ghana http://ugspace.ug.edu.gh P e a n u t s Melon seeds Tiger nuts Coconut Sort Sort Sort Shell/Pare Wash Wash Wash Wash Roast at Roast at Roast at 120OC for 120OC for 120OC for 40 minutes 20 minutes 15 minutes Roast at Roast at Roast at 120OC for 40 120OC for 40 120OC for 40 Blend with mDiensh ull Blmeninds with Blemninds with water. Coconut: water. Melon water. Tiger water -1:3 seed: water -1:5 nut: water -1:5 Blend with Roast at Roast at Roast at O water. Peanut: 120 C for 40 120 O C for 40 Bromelain 120 OC for 40 mins water -1:4 mins (0.005%v/v/6 mins 0 min) at 28oC Sieve Sieve Sieve S i e v e Filtrate Filtrate Filtrate Filtrate Buffer, Buffer, Buffer, stabilizer, stabilizer stabilizer suga , sugar r , sugar Homogenise H omogenise Homogenise Homogenise Peanut milk Melon seed milk Tiger nuts milk Coconut milk Combine according to ratios Pasteurize at o generated by mixtures design 80 C for 20 Bottle ( T a b l e 3 . 1 ) m i n u t e s Plant milk beverages Figure 3.5 1 Process flow diagram for plant milk processing 35 University of Ghana http://ugspace.ug.edu.gh 3.5 Characterization of optimized beverages Product characterization was done in three ways; 1. Physicochemical analyses to determine the physicochemical characteristics of only the 3-blend prototypes, 2. Relative Preference Mapping (RPM) to determine which of the 3-blend beverages were acceptable to consumers compared with commercial plant and dairy milk beverages and 3. Consumer acceptance test of all 19 formulation prototypes using the BIB design to determine which of the main ingredients in the 3- blend plant milk prototypes influenced consumer liking. 3.6 Physicochemical characterization The physicochemical characteristics of the 3-blend plant milk beverages, the commercial plant and dairy milk beverages were determined as follows: 3.6.1 pH and Titratable Acidity The pH of the products was measured with a pH meter (Model HM-30S, TOA Instruments, Japan). Titratable acidity was determined using the method described by Aidoo et al. (2010); for each beverage, 10 ml was titrated with 0.1N NaOH with 1% phenolphthalein as indicator. Acidity was calculated as lactic acid (% m/v). The determinations were carried out in triplicate, means and standard deviations are reported. 36 University of Ghana http://ugspace.ug.edu.gh 3.6.2 Color The color of the beverages was measured as described by Aidoo et al. (2010), measurements were based on the L*a*b* color system and was determined using a Minolta Chroma Meter (Data Processor DP-301, for Chroma Meter CR-300 series). The color difference (ΔE) was determined using the following formula: [ΔE = (L*- L) 2 + (a*- a) 2 + (b*-b) 2]1/2. The standard white tile to which the samples were compared had the following color indices: (L* = 97.95, a* = -0.12, b* = + 1.64). The determinations were carried out in triplicate, means and standard deviations are reported. 3.6.3 Viscosity The viscosity, flow index and consistency were measured using the Haake Viscotester D (399- 0401, Thermo Scientific™, Spain) based on the method described by Bastioglu et al (2016). For each beverage, 100ml was tempered to 20oC prior to evaluation. The shear rate of the Viscotester was set at 12, 30, and 60 s-1 and determinations were carried out in duplicate. Means and standard deviations are reported 3.6.4 Proximate composition Total moisture was determined as described by Pinelli et al. (2015). For each beverage 20ml was dried in a hot air oven at 105oC till constant mass. The nitrogen content (N) of the samples was determined using the Kjeldhal method as described by AOAC (2005) method 991.20. The total protein was calculated by multiplying N (the total nitrogen) with 6.25. Total solids were determined according to standard procedures as described by AOAC (2005) method 990.20. The 37 University of Ghana http://ugspace.ug.edu.gh total fat content was determined using the Rose-Gottlieb Method AOAC (2005) method 905.02. The total ash was determined by heating 2g of the sample in a furnace at 600oC for 6 hours as described by AOAC (2005) method 935.45. Total carbohydrate content was determined by difference as described by Pinelli et al. (2015); Crude Carbohydrates = 100% - (%protein + %moisture + %crude fat + % crude ash). 3.7 Sensory characterization of optimized products 3.8 Comparing the sensory acceptability of 3-blend samples with commercial dairy and plant milk beverages using Relative Preference Mapping (RPM) RPM was designed to be used in a social setting, however for the purpose of this study, it was important to test that comparable results are obtained in both a social setting and laboratory setting (where sensory tests are normally done). As such RPM was done in both a social and laboratory setting to determine reliability of results. Trials of the RPM were done in both the social and laboratory setting. This was to make sure that the RPM will give reliable results before the actual consumer testing of the 3-blend products. The laboratory trial is described in Appendix 7. 3.8.1 Test locations For the social setting, product fairs were used. Table 8.1 in Appendix 8 shows the different events serving as test location for RPM tests in the social setting and the characteristics of those locations that qualified them as a social event context. The laboratory test location was the Sensory Evaluation Laboratory at Department of Nutrition and Food Science at University of 38 University of Ghana http://ugspace.ug.edu.gh Ghana. The laboratory is fitted with 8 individually partitioned booths, a discussion area and preparation area. It has white light and is noise and odor free. Researchers have minimal contact with assessors and samples are served through a hatch from the preparation area. The environment is temperature controlled by air conditioning and the temperature is maintained at 25oC ± 2. See Figures 8.1 to 8.3 in Appendix 8 for pictures of the test at both locations. 3.8.1 Participants In the social setting, an announcement was made to inform people about the test. For the laboratory tests, consumers were recruited from the University of Ghana Campus. In both settings, recruitment was by one-on-one interviews using a recruitment questionnaire (Recruitment questionnaire for this test is presented in Appendix 2). The selection criteria were as follows; potential participants were to be consumers of plant milk and were not to be allergic or intolerant to milk and milk products, peanuts, melon seeds, soybeans, tiger nuts or any of the ingredients used in the beverages. Participants who qualified were selected for the test. A total of 90 plant milk consumers aged between 17-54 years were used each for the laboratory and social RPM (making a total of 180 assessors). Assessors signed a consent form and demographic data was collected from them before they started the tests. Guardians of assessors under 18 years signed a parental assent form. At the end of the study, assessors were given a small token as compensation for their time. 39 University of Ghana http://ugspace.ug.edu.gh 3.8.3 Products The 3-blend beverages were selected from the runs generated by the mixtures design shown in Table 3.2. The products were microbiologically tested to ensure they were safe for consumption before they were used for the consumer tests. A commercial dairy milk beverage was used as reference and the leading brand of plant milk currently on the market was included in the product set to estimate how the current blends compare with a commercial product. The full product set is shown in Table 3.3. Table 3 3 Products used for RPM Product Number Product Code Product Details 1 Melon seeds milk (37.5%), Coconut milk (25%) and Peanuts C milk (37.5) 2 Melon seeds milk (25%), Coconut milk (50%) and tiger nuts D milk (25%) 3 I Coconut milk (50%), Tiger nuts milk (25%) and Peanuts milk (25%) 4 Melon seeds milk (25%), Coconut milk (50%) and Peanuts N milk (25%) 5 Melon seeds milk (37.5%), Coconut milk (25%) and tiger P nuts milk (37.5%) 6 Coconut milk (25%) Peanuts milk (37.5%) and Tiger nuts Q milk (37.5%) 7 V Vitamilk (Original) 8 (BC) Even UHT Full Cream Milk (with 0.04% w/v sugar Reference added) The 3-blend beverages were processed as outlined in section 3.4.2. Vitamilk and Even UHT milk were bought from the supermarket. 0.04% w/v of cane sugar was added to the UHT milk to sweeten it (the same quantity of sugar used to sweeten the plant milk). Products were stored at 40 University of Ghana http://ugspace.ug.edu.gh 4oC. For the RPM at the social events, products were transported to the events in thermos flasks with ice packs in order to maintain the temperature of the products. 10ml of products were served in 80ml disposable clear cups labeled with three-digit product codes, whilst 20ml of the reference was served for participants to compare with all products. The products were served as and when they were needed. For the laboratory-based RPM, 10ml of products were served in 20ml plastic cups whilst 20ml of reference was provided. The products were pre-poured and kept in the refrigerator and assessors were served from the refrigerator in order to maintain the product temperature of 8oC-10oC. All products were served in a monadic sequential order according to a balanced design, using the William’s design in Compusense Saas (Compusense® Inc. Guelph, Ontario, Canada). 3.8.4 General test procedure The map scale used for the RPM was a T-Map scale as shown in Figure 3.8.1. It has two axes, the y-axis which is the liking axis at the center of it is the reference and the x-axis (the difference axis) which is perpendicular to the y-axis and starts from the center at the reference. If a product code is placed above the reference on the y-axis, it means that the product is liked more than the reference; the further away a product is placed from the reference, the more the product is liked more than the reference. For product codes placed below the reference on the y-axis, the further away the product code is placed from the reference, the more it is disliked compared to the reference. The x-axis is the difference axis; the closer a product code is placed to the reference on the x-axis, the more similar the product is perceived to be to the reference, whilst the further away a product is placed from the reference, the more different the product is perceived to be from the reference. 41 University of Ghana http://ugspace.ug.edu.gh In the social RPM, the T-Map was printed on an A0 sheet, which was set up using a flip chart. Assessors were given a strip of colored dots with the different product codes: they were instructed to place the product codes on the T-Map scale based on their assessment. In the laboratory, the test was set up using Compusense Saas (Compusense® Inc. Guelph, Ontario, Canada), assessors did their evaluations on computers in individual partitioned booths. Research assistants who assisted in the test were briefed about the test procedures before the beginning of the test and were told to direct all difficult questions to the lead researcher. Assessors were given a verbal explanation of the test procedure after which they were given the reference product to evaluate first and then were given the plant milk products. They were instructed to simultaneously evaluate how different and how much they like or dislike the products compared to the reference. After each product, they were made to rinse their mouths with water. Figure 3.8 1 T-Map scale 42 University of Ghana http://ugspace.ug.edu.gh 3.9 Consumer acceptance test of all 19 formulation prototypes using the BIB design to determine which of the main ingredients in the 3-blend plant milk prototypes influenced consumer liking A consumer test using the balanced incomplete block design was used to understand the ingredients driving liking for the 3-blend beverages 3.9.1 Products All 19 formulations generated from the mixtures design were used for this consumer test. Refer to Table 3.2 for the formulations. These products were microbiologically tested to ensure their safety before being served to assessors. 3.9.2 Test Location The consumer test was carried out at the Sensory Evaluation Laboratory at the Department of Nutrition and Food Science, University of Ghana. 3.9.3 Participants Assessors were recruited from the University of Ghana campus. Potential assessors were approached and told about the test and asked if they would be willing to participate, when they gave their consent, they were screened to see if they met the selection criteria (this was done using the screener in the recruitment questionnaire for the test, recruitment questionnaire is 43 University of Ghana http://ugspace.ug.edu.gh presented in Appendix 2). Plant milk consumers who were not allergic or intolerant to milk and milk products, peanuts, melon seeds, soybeans, tiger nuts or any of the ingredients used in the beverages were selected. After screening, nineteen (19) plant milk consumers aged between 18- 34 years were selected for the test. They signed a consent form and demographic information was collected from them. They were given a small token as compensation for their time. 3.9.4 General Test procedure The scale used for this consumer test was the traditional 9- point hedonic scale with descriptors and numbers. With a score of 1 meaning: dislike extremely and a score of 9 meaning like extremely. This was a paper-based test (ballot sheet is presented in Appendix 6) Each assessor evaluated 9 products as randomized using the SAS software. 15ml of each product was served in 20ml plastic cups. All products were served at 8oC-10 oC in a monadic sequential order as randomized using the SAS software according to the BIB design. Assessors were asked to evaluate the products based on their overall liking of the product first. After that they proceeded to evaluate how much they liked the appearance, flavor, mouth feel, consistency and after taste. After evaluating each product, assessors were asked to write what they liked and disliked about each product. They rinsed their mouths with water before proceeding to the next product. 44 University of Ghana http://ugspace.ug.edu.gh 3.10 Data Analysis 3.10.1 Focus group discussion Audio data was transcribed and compared with hand written notes. Thematic analysis was used to analyze the data by grouping the information under themes and sub-themes. After that conclusions and inferences were made from the analyzed data. Facial expressions and other non- verbal cues were taken into consideration in interpreting the data. 3.10.2 Physicochemical characteristics These were analyzed using one-way analysis of variance using XL-Stat (Addinsoft 40, rue Damrémont 75018, Paris). Post- hoc analysis was based on Fischer’s LSD. 3.10.3 Consumer test using Relative Preference Mapping The difference and liking axes were each 50 cm long with intervals of 1 cm. The scores were derived by reading the mark made by each assessor from the end of each axis. For the difference axis, the end R was taken as the starting point and as the mark moved away from the R, it was considered more different from the reference. For the liking axis, the starting point was at the bottom of the axis and increased as the data points moved up. Lower values meant that the product was disliked or was similar compared to the reference and high values meant that the product was liked and was different compared to the reference. Data from the liking axis from the social and laboratory setting were analyzed using a two-way analysis of variance (ANOVA). Individual product maps created by consumers were condensed into a consensus product map 45 University of Ghana http://ugspace.ug.edu.gh using Generalized Procrustes Analysis (GPA) method. The region of innovation was highlighted as the space between the liking and difference axes on the consensus map. All analyses were carried out using XL-Stats (Addinsoft 40, rue Damrémont 75018, Paris). 3.10.4 Data analysis of consumer acceptance test using BIB design The consumer test using the balanced incomplete block design gave consumer liking data and consumer comments. Analysis of consumer liking data Regression analysis was carried out on the consumer liking data and the 19 formulations using Minitab 17. Cox response plots were generated for the overall liking, appearance, flavor, mouth feel, consistency and aftertaste. These show the effect of the various ingredients on the liking scores for the overall liking, appearance, flavor, mouth feel, consistency and aftertaste. Overlaid contour plots were also generated for the dependent variables (i.e. acceptance scores). To generate the overlaid contour plot, an optimum region for mean acceptability scores for all sensory attributes was set from 7 (like moderately) to 9 (Like extremely). The optimum region is the area where the criteria set for acceptability is met (Aidoo et al., 2012). Product liking scores were analyzed using one-way analysis of variance using Minitab 17. Post-hoc analysis was based on Fischer’s LSD. 46 University of Ghana http://ugspace.ug.edu.gh Consumer Comments Consumer comments were analyzed as described by Symoneaux et al. (2012). Consumer comments were either “like” comments prefixed with L_ and “dislike” comments prefixed with D_. To ensure that only descriptive terms were used, affective terms, examples “good” and “nice” were removed. Ambiguous words were also removed to avoid misinterpretation. Total likes and dislikes per product were counted and a contingency table of the like and dislike comments were created. The contingency table was analyzed by correspondence analysis using XL-Stats (Addinsoft 40, rue Damrémont 75018, Paris) to create a product map for products. 47 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR 4.0 RESULTS AND DISCUSSION 4.1 Focus group discussion During the focus group discussions, five main themes emerged: 1. Knowledge about plant milk, 2. Consumption of plant and dairy milk, 3. Marketing of plant milk, 4. Usage of plant milk and 5. Raw material for preparation of plant milk. 4.1.1 Knowledge about plant milk The plant milks the participants were familiar with included soymilk, tiger nut milk, coconut milk, almond milk and rice milk. Some participants indicated that soymilk was popular because it does not have a detrimental effect on the environment when cultivated, while others indicated that tiger nut milk was more popular than soy milk. But all the participants agreed later on in the discussion that soymilk was more popular. 4.1.2 Consumption of plant and dairy milk Plant milk consumed by participants included rice milk, soymilk, coconut milk, almond milk and tiger nut milk. Soy milk (Vitamilk) was the most commonly consumed plant milk. A participant mentioned that he had consumed soymilk ice cream which contained coconut milk. The frequency of plant milk consumption varied amongst the participants. Some participants consumed plant milk every day, twice a week, three times a week or every other week. Some 48 University of Ghana http://ugspace.ug.edu.gh consumed it as and when they wanted to consume plant milk. A participant made the following comment: “I take it as and when I want to” Some participants complained that because of the cost, they did not consume plant milk as often as they wanted to. Some of the comments by participants were: “Because of the cost, sometimes I consume it once a week” “Occasionally, I cannot say if it’s once or twice, just occasional because of the price”. “Occasional, when there is the need, depending on the pocket and feeling” All participants indicated that they consumed dairy milk, but they did not consume fresh pasteurized milk. The frequency of dairy milk consumption also varied. Some consumed dairy milk every day and some once a week. The following were comments made by participants: “I consume dairy milk every day, as part of my breakfast, I drink carnation” “Every day, I drink ideal milk; I add it to my tea” “I drink dairy milk once a week with gari” Some participants indicated that they consumed plant milk as a snack with biscuits, bread or pastries while others indicated they consume it as a beverage. A participant consumed plant milk as a precursor to exercise. A participant stated: “I take it when I am going to play football” 49 University of Ghana http://ugspace.ug.edu.gh Others consumed plant milk as a nutritional drink and when tired as the following comments indicate: “I take Vitamilk as a nutritional drink; I prefer it to fizzy drinks. I give it to my daughter as a snack to school” “I take it when I am tired in the evening” Some participants consume plant milk as a whole meal or use it as an ingredient in recipes as the following comments indicate: “I use almond milk for pancakes” “I use plant milk for yogurt and vanilla ice cream and add milk to it” Some participants indicated that some people drink plant milk because they are lactose intolerant. Participants gave various reasons why they consumed plant milk, some indicated that they liked the beany flavor of plant milk while others indicated that plant milk tasted fresher compared to dairy milk. Other reasons included knowledge about the perceived nutritional and health benefits: they perceived plant milk as safer, contained less sugar and had no cholesterol. Some of the statements made are as follows: “Because I like the beany flavor” “Feels fresh as compared to dairy milk, I like the beany flavor” “Before my university education, I took plant milk just as a beverage with no reason but after university, I know of cholesterol and other health benefits so consume it more”. “It contains less sugar” 50 University of Ghana http://ugspace.ug.edu.gh “I just like it” “I consume plant milk because it is low in cholesterol” “Parents think it is nutritional and healthy, so they buy it as motivation for good behavior”. “Because of nutritional and health issues; been drinking it from infancy as a training from parents” During the discussion, participants got to know that some producers add dairy milk to some plant milk brands and many of them were not happy with this information. 4.1.3 Marketing of plant milk According to some participants, they buy plant milk from retail shops, supermarkets, convenient shops, local markets, street vendors, from petty traders, and anywhere plant milk was available. Some participants also make the plant milk themselves. Some verbal comments include: “I purchase them from shops” “Almond milk is hard to get. Will go to the local market to get tiger nut milk. You get almond and coconut milk from high end shops” A participant indicated that she makes plant milk herself; “I make the coconut milk myself” According to participants, there is demand for plant milk because plant milk in general and Vitamilk to be specific is doing well on the market. They also indicated that a reason for the 51 University of Ghana http://ugspace.ug.edu.gh demand for plant milk could be because it helps lactose intolerant people. Participants suggested that children should be the target of plant milk marketing as in these statements: “Yes, children like Vitamilk and it will depend on how it is marketed and they do not see difference between plant and animal milk so they like it”. “Overtime, with children as the target, there will be a big market. It will depend on what you are exposed to. There are not many almond trees planted on a large scale. Almond milks are expensive and only in supermarkets. Apart from soymilk, other plant milks are not common” A participant indicated that the cost of plant milk could be an impediment to its sales. But this was debunked by the other participants because most of the time plant milk is sold in one litre tetrapaks and when compared by volume to dairy milk, plant milk is not more expensive. Participants also suggested that the 1 litre packs are too much and the volumes of the packages should be reduced as in the following statement: “The tetrapak is too big in size. Vitamilk is expensive. It would be good to sell small sizes with lower price so that everyone can buy” Participants suggested that plant milk be marketed as snacks, milk substitutes, beverages and as a creamer. They advised marketers to advertise the nutritional benefits of plant milk. They all agreed that the target market for plant milk should be everybody, children and vegetarians. 4.1.4. Usage of plant milk Some participants indicated that because of the beany taste of plant milk they will not use it as a creamer. They indicated that they might have to use more plant milk because of the taste of the 52 University of Ghana http://ugspace.ug.edu.gh plant milk. Some also indicated that they have not considered it for use as a creamer. A participant made the following statement: “I will have to consider the taste, so will have to add more”. A participant indicated that she uses plant ingredients in place of dairy milk to prepare porridge for her children as indicated as follows: “I dehull cowpea and soybeans, add it to corn dough and cook; it comes out as if you have added milk”. Other participants indicated that plant milk can also be used to replace milk in yogurt preparation. 4.1.5 Raw materials for preparation of plant milk Participants indicated that tiger nut milk is very popular on the local market and should be commercialized. They also suggested plant milks made from maize, coconut, rice, millet, cowpea and peanut should also be commercialized. As in the following statements: “Tiger nut is nutritious” “Commercialize tiger nuts milk” Participants liked the color and taste of coconut milk. The following are comments made when coconut milk was tasted by participants: “I like the color”. 53 University of Ghana http://ugspace.ug.edu.gh “It looks like cow milk”. “It is whiter than cow milk”. “I like the appearance; the color is natural”. “It looks like the milk you get when you crush palm nut” “I like the color, texture, coconut flavor”. “The appearance is not bad, but there is too much chaff”. “I like the smell”, it can be used to make biscuits”. “I taste coconut”. “It is not too sugary”. “My mother uses coconut milk to make rice water”. “It leaves an oily film on your tongue”. “I wish it was thicker”. “It is too light”. “It is okay as it is”. “I am happy with how it is”. Two participants indicated that they will use this coconut milk as a snack. One person indicated she will use it as part of a cocktail and one person for her hair. All participants suggested that it should not contain particles and it should be thicker. 54 University of Ghana http://ugspace.ug.edu.gh “I would use it for gari without water”. “Will rather take this one as snack, should be thickened so it can be used as a snack”. “It’s overpowering so it will not work as a creamer”. Participants indicated that peanuts are not suitable to be used for plant milk because peanuts have high oil content. They indicated that peanut is known for being used for groundnut paste and that there are other uses for it. During the tasting session, participants liked the color and texture of peanut milk but did not like the taste and aroma. The following are comments made when peanut milk was tasted by participants: “The color is attractive” “I like the color, like peanut”. “Off-white, raw peanut”. “Has oil on top of it”. “It tastes like fresh peanut” “The aroma is bad, the taste is bad, but I like the appearance” “Too strong fresh unroasted peanut aroma” “Smells like neem, it may contain antimalarial” “I do not like the taste and aroma” “Has natural smell” “I do not like the flavor and aftertaste” “It is nauseating, I have never seen or consumed it before, there is nothing to like about it. I do not like the taste, aroma and it’s too watery” 55 University of Ghana http://ugspace.ug.edu.gh “No sugar, too strong peanut aroma” “It is creamy” “It is smooth They made the suggestions below: “Try roasting the groundnuts” “Add sugar to the milk”, sugar can be added and peanut roasted to make it better”. “If there is a bit of sugar, it will be okay”. “Plant smell is too much, too strong and doesn’t taste so sweet. If it is sweet, people won’t notice”. “It has to be flavored with vanilla and strawberry”. “Roast the peanut or try boiling it”. Participants did not think melon seeds would be a good ingredient for plant milk. The following are comments from participants: “It does not sound appetizing” “It does not sound normal” “You never know” Participants indicated that melon seed milk had roasted peanut aroma. Participants liked the color and the aroma but did not like the taste and aftertaste. All participants complained that it had a bitter aftertaste and that it was oily and non-homogeneous. “I like the aroma”. “It is not nice”. 56 University of Ghana http://ugspace.ug.edu.gh “It tastes like wrewre soup and I don’t like wrewre soup”. “We have not tasted this type of milk before”. “It tastes more like egusi, I like the color and creaminess”. The sugar content is fine”. “I want more sugar”. During the focus group discussion participants liked the commercial soymilk beverages (Vitamilk and Lactasoy), the only laboratory prepared plant milk participants liked during the focus group discussions was coconut milk, as such, it was decided that coconut milk should be in all the formulations, so the minimum quantity of coconut milk was kept at 25% in all the formulations. Consumers stated that tiger nuts were nutritious and tiger nut milk was popular and should be commercialized, because of this, tiger nut was added to the other raw materials (melon seeds, peanuts and coconuts) with the aim of increasing the acceptance of the optimized beverages as tiger nuts milk was already popular and accepted. Even though participants did not like the taste of melon seeds milk, there was no recommendation to improve it. It was kept as a raw material because it was thought that the other raw materials with better flavor will mask any off-flavors melon seeds milk might impact in the 3-blend formulations. During the focus group discussions, participants described the peanuts milk as tasting of raw peanuts. Because of this, participants in the focus group discussions recommended that the peanuts be roasted before blending with water. 57 University of Ghana http://ugspace.ug.edu.gh 4.2 Characterization of optimized products 4.3 Physicochemical characterization 4.3.1 Proximate composition of 3-blend, commercial plant milk and dairy beverages Table 4.1 shows the proximate composition of the 3-blend plant milk, commercial plant milk and dairy beverages. Table 4 1 Proximate composition of beverages Composition (%) Product Moisture Protein Fat Ash Carbohydrates Total Solids C 85.26±0.01f 3.51±0.02 ab 6.30±0.42 a 0.25±0.01f 4.68±0.01g 14.74±0.01a D 88.22±0.13a 1.89±0.24 de 5.11±0.20 b 0.49±0.01bc 4.28±0.01h 11.78±0.13f I 87.36±0.23c 1.65±0.06 e 5.00±0.35 bc 0.34±0.00de 5.65±0.01d 12.64±0.23d N 85.89±0.22e 2.16±0.39 cd 6.54±0.04 a 0.31±0.02ef 5.09±0.01f 14.23±0.22b P 87.19±0.02c 2.31±0.21 c 4.09±0.05 d 0.51±0.07b 5.90±0.03c 12.81±0.02d Q 85.77±0.01e 1.90±0.09 de 4.71±0.08 c 0.36±0.02de 7.26±0.01a 14.24±0.01b V 86.89±0.11d 3.30±0.04 b 3.36±0.00 e 0.41±0.01cd 6.04±0.01b 13.18±0.11c BC 87.84±0.01b 3.76±0.00 a 2.49±0.22 f 0.71±0.02a 5.23±0.06e 12.17±0.02e Legend: C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q (25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk); V (Vitamilk) and BC (Even UHT full cream milk with added sugar). The total solids content is an indication of the level of nutrients in a product. Product C (M37.5%, C25%, P37.5 %) had the highest total solids content and this was significantly different from that of product N (M25%, C50%, P25 %) which had the second highest total solids. Between the two products (C and N), product C’s higher content of melon seeds milk and peanuts milk could have accounted for the higher total solids content of product C (M37.5%, C25%, P37.5 %). Product D (M25%, C50%, T25 %) had the lowest total solids content of 11.78%. This product contained a high quantity of coconut milk (50%) which could account for the low total solids content. Odoom (2018) 58 University of Ghana http://ugspace.ug.edu.gh found that plant milk beverages with high amounts of coconut milk had low total solids content as they had low amounts of proteins, carbohydrates and ash, while products with high amounts of peanuts and melon seeds milk had high total solids content as they contained high amounts of carbohydrates, proteins and ash. Oyeike and Acheru (2002) also found coconut seeds to have the lowest amount of ash, proteins and carbohydrates when they compared the proximate composition of coconut seeds with peanuts, melon seeds and other oil seeds. Five out of the six experimental plant milk products had total solids higher than the dairy milk sample, while three, C (M37.5%, C25%, P37.5%), N (M25%, C50%, P25%) and Q (T37.5%, C25%, P37.5%) had higher total solids compared to the commercial plant milk product (Vitamilk). Product BC (the dairy milk product) had the highest protein content of 3.76 ± 0.00%. This protein content was not statistically different from the protein content of product C (M37.5%, C25%, P37.5%) which had a protein content of 3.51±0.02%. Product C (M37.5%, C25%, P37.5%) had higher protein content than the commercial plant milk product (Vitamilk). Product C (M37.5%, C25%, P37.5%) contained a high content of melon seeds and peanuts milk which are both rich in protein. Peanuts have protein content of 23.68% (Swati et al., 2016) while melon seeds have protein content of 25.4% (Oyeike and Acheru, 2002) which could account for the high protein content of this product. Product I (C50%, T25%, P25%) had the lowest protein content of 1.65 ±0.06% which could be because of its high coconut milk content. Coconut milk has a protein content of 2-4% (Tansakal and Chaisawang, 2006). The plant milk products had higher fat contents than the dairy milk product (BC) and the commercial plant milk product V (Vitamilk). This is because the raw materials used for the plant milk were all high in fat, having fat contents greater that (40%). Product N (M25%, C50%, P25 %) had the highest fat content which could be because of its high coconut milk content of 50%. The dairy milk product (BC) had the highest ash content of 0.71±0.02%. The ash content of dairy milk is 59 University of Ghana http://ugspace.ug.edu.gh generally higher than plant milk and some plant milks have to be fortified with minerals and vitamins. Product P (M37.5%, C25%, T37.5 %) had the highest ash content amongst the plant milk products (0.51±0.07%). The product with the lowest ash content was product C (M37.5%, C25%, P37.5%), with an ash content of 0.25±0.01%. All products containing peanut milk had ash contents below 0.40%. This was also the trend when Aidoo et al. (2010) combined peanuts and cowpea to replace skimmed milk powder in chocolate. The plant milk powders with high amounts of peanut had low ash content. Product Q (T37.5%, C25%, P37.5%) had the highest carbohydrate content which may be due to the high carbohydrate content of tiger nuts (43.3%) which is mainly made up of starch (Sánchez-Zapata et al., 2012). Product D (M25%, C50%, T25 %) had the lowest carbohydrate content of 4.28±0.01%. 4.3.2 Color of 3-blend, commercial plant milk and dairy beverages Table 4.2 shows the color indices of the plant milk products. Table 4 2 Color indices of beverages Product L a b ΔE C 83.13±0.32c 0.86±0.02e 13.74±0.06d 19.07±0.06f D 80.06±0.77d 0.58±0.02f 13.21±0.05f 21.32±0.01d I 79.32±0.05e 0.98±0.05c 13.62±002e 22.17±0.01c N 85.54±0.50b 0.52±0.04g 12.74±0.05g 16.64±0.01g P 75.91±0.13g 0.91±0.01d 15.36±0.04b 25.93.±0.01a Q 78.22±0.13f 1.87±0.03a 14.61±0.01c 23.69±0.01b V 85.31±0.02b 1.41±0.01b 16.25±0.04a 19.36±0.01e BC 95.23±0.10a -0.93±0.01h 13.66±0.05e 12.01±0.01h Legend: C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q (25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk); V(Vitamilk) and BC (Even UHT full cream milk with added sugar). 60 University of Ghana http://ugspace.ug.edu.gh The plant milk products generally creamy white in color. All the 3-blend beverages were significantly different from the each other in terms of color. The L value measures the lightness of the product, the higher the L value, the whiter the product. Product BC, which is the dairy milk product had the highest L value and thus was the lightest. The products with high amounts of coconut milk and melon seeds milk were lighter, whilst those with tiger nuts milk were darker. Those with peanut milk had medium lightness due to the color of coconut, peanut, tiger nuts and melon seed milk as depicted in Figure 4.3.1. Products N (M25%, C50%, P25 %) and V (Vitamilk) were statistically not different from each other and were the lightest amongst the plant milk products. The high percentage of coconut milk in product N (M25%, C50%, P25 %) might have imparted a lighter color to it. The two darkest products, P (M37.5%, C25%, T37.5 %) and Q (T37.5%, C25%, P37.5 %) with L values of 75.91±0.13 and 78.22±0.13 had high percentages of tiger nut milk in their formulation which could account for the darkness of the products. The b value when positive signifies a high yellowness of the product. The high positive b values indicate that the products are yellowish or have yellowish tinges. High positive a values indicate redness, whilst low or negative a values indicate greenness. The low a values of the products indicate that the products have very low amounts of redness in them and are more greenish. The ΔE values indicate the color difference between the products and a standard white tile. Sample P (M37.5%, C25%, T37.5 %), the darkest sample had the highest ΔE of 25.93±0.01 whilst sample BC (the dairy milk product) had the lowest ΔE of 12.01±0.01. 61 University of Ghana http://ugspace.ug.edu.gh Figure 4.3 1 Color of the different plant milks used in the formulations 4.3.3 pH and titratable acidity of 3-blend, commercial plant milk and dairy beverages Table 4.3 shows the pH and titratable acidity of the beverages. Table 4 3 pH and titratable acidity of beverages Product pH Titratable acidity C 6.85±0.02ab 2.05±0.00a D 6.83±0.01ab 1.39±0.01c I 6.84±0.05ab 0.83±0.08ef N 6.76±0.04cd 1.78±0.08b P 6.74±0.00d 1.78±0.03b Q 6.87±0.01a 0.89±0.01e V 6.81±0.05bc 0.75±0.01f BC 6.85±0.00ab 1.17±0.06a Legend: C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q(25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk); V(Vitamilk) and BC (Even UHT full cream milk with added sugar). 62 University of Ghana http://ugspace.ug.edu.gh Titratable acidity is an indication of the total acid content of the various products. Product C (M37.5%, C25%, P37.5 %) had the highest titratable acidity of 2.05±0.00, whilst V (Vitamilk) had the lowest titratable acidity of 0.75±0.01. The melon seed milk containing formulations have very high titratable acidity, the melon seed milk was treated with a protease (Bromelain) to reduce the viscosity and this could account for the high titratable acidity. This is because protein hydrolysis leads to the release of H+ ions in the reaction medium and will need more OH- ions to neutralize them thus leading to increased titratable acidity (Aluko, 2018). 4.3.4 Rheology of 3-blend, commercial plant milk and dairy beverages Table 4.4 shows the rheological properties of the beverages. Table 4 4 Rheological parameters of beverages Product Repetitions Apparent Viscosity Flow behavior R2 (Centipoise) index C 2 1527.05±4.03d 0.5875±0.00e 0.9994 D 2 1622.60±4.95b 0.6340±0.00bc 0.9987 I 2 1814.85±2.47a 0.6415±0.00b 0.9998 N 2 1340.85±1.77e 0.6295±0.00c 0.9991 P 2 1163.55±7.42f 0.6145±0.00d 0.9960 Q 2 1605±2.33c 0.5970±0.01e 0.9977 V 2 118.25±2.39g 0.6085±0.01d 0.9974 BC 2 49.32±0.05h 0.777±0.00a 0.9961 Legend: C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q(25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk); V(Vitamilk) and BC (Even UHT full cream milk with added sugar). The apparent viscosities of the products were all significantly different from each other. The 3- blend products were highly viscous because of the high amount of stabilizer added to them (0.2%). Another reason that could account for the higher viscosities is the presence of coconut 63 University of Ghana http://ugspace.ug.edu.gh milk. Coconut milk has higher viscosity because its dispersed phase particles are larger and this results in higher viscosities (Aboufazli et al., 2014). The product with the highest viscosity was product I (T25%, C50%, P25 %), with a viscosity of 1814.85±2.47. Generally, products containing melon seed milk had lower viscosities compared with the products without melon seed milk, only product D (M25%, C50%, T25 %) was the exception as it had the second highest viscosity of 1622.60±4.95. Products I (T25%, C50%, P25 %,) and D (M25%, C50%, T25 %) both contain 50% coconut milk, the high quantity of coconut milk in product D (M25%, C50%, T25 %) could explain its high viscosity. What could account for the generally lower viscosities of the products containing melon seeds could be the treatment of the melon seeds milk with a protease (Bromelain) which denatured the proteins, reduced the gelling during heating and led to lower viscosities. Product V (Vitamilk) and Product BC (the dairy milk product) were the lightest products. All the products were shear-thinning, their viscosities decreased with increasing shear rate (Tangsuphoom and Coupland, 2005). The flow behavior index gives an indication of the pseudo plasticity of the product (Aboufazli et al., 2014). Sample BC (dairy milk product) had the highest flow behavior index and was significantly different from the plant milk products. Amongst the 3-blend products, products C (M37.5%, C25%, P37.5 %) and Q (T37.5%, C25%, P37.5 %) had the lowest flow behavior index and they were not statistically different. This could be because of the low amount of coconut milk (25%) in these products. Product P (M37.5%, C25%, T37.5 %) also had a low flow behavior index because it contained a low amount of coconut milk. This trend was also reported by Aboufazli et al. (2014). They studied the rheological properties of ice cream in which skimmed milk powder had been partially or completely replaced by either soymilk or coconut milk. Those ice creams with a lower percentage of coconut milk had low flow behavior index 64 University of Ghana http://ugspace.ug.edu.gh values. Figure 4.3.2 shows the effect of increasing shear rate on the apparent viscosity of the products. 400 350 300 250 200 150 100 50 0 0 10 20 30 40 50 60 70 Shear rate C D I N P Q V BC Power (C) Power (D) Power (I) Power (N) Power (P) Power (Q) Power (V) Linear (BC) Figure 4.3 2 Effect of shear rate on apparent viscosity of products 4.4 Sensory Characterization of optimized beverages 4.5 Comparing sensory acceptability of optimized 3-blend beverages with commercial dairy and plant milk beverages using RPM 4.5.1 Comparison of mean liking scores of 3-blend beverages with commercial products Figure 4.5.1 shows the liking scores obtained for the RPM in both the social and laboratory settings 65 Apparent Viscosity University of Ghana http://ugspace.ug.edu.gh 50 45 A A 40 B B B 35 B 30 CD C CDE 25 DE E 20 F F FG 15 G G Lab 10 Social 5 0 Samples Figure 4.5 1 Comparison of Mean liking scores (Y-Axis) between 3-blend products and commercial samples Error bars are standard errors of the mean. Data labels represent Fishers LSD Post hoc analysis of the means. Samples with no letters in common are statistically different at the 95% confidence level. Legend: C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q(25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk); V(Vitamilk) and BC (Even UHT full cream milk with added sugar). A commercial dairy milk beverage was used as reference for the consumer test using RPM. An earlier study by Diarra et al. (2005) suggested that the more similar the sensory attributes of plant milk is to dairy milk, the more it will be accepted. In order to measure consumer acceptance, the 3-blend beverages needed to be compared to a commercial plant milk beverage that was already accepted by consumers. Vitamilk, a soymilk product was chosen for comparison with the experimental plant milk products as it was considered the most popular and well-liked plant milk beverage in Ghana by participants during the focus group discussions. The longest bar on Figure 66 Mean Liking Scores University of Ghana http://ugspace.ug.edu.gh 4.5.1 represents the most liked product while the shortest bar is the least liked. Even though the liking scores for the products were higher in the social setting, the rank order for liking scores for the social and laboratory were similar, except that, in the social setting, product C(M37.5%, C25%, P37.5 %) was the least liked whilst in the laboratory setting products C (M37.5%, C25%, P37.5 %), N(M37.5%, C25%, P37.5 %) and P(M25%, C50%, P25 %) were the least liked. It has been found by various researchers that some foods are rated higher when tested in more “natural” setting compared to the laboratory setting, but generally the rank orders have been the same for the different settings (De Graaf et al., 2005; Boutrolle, et al., 2007). The liking score for product BC (the dairy milk product) was not statistically different from product V (Vitamilk). Both products were the most liked. All the 3-blend products were less liked than the products BC (the dairy milk product) and V (Vitamilk). Products I (T25%, C50%, P25%) and Q (T37.5%, C25%, P37.5%) were the most liked amongst the 3-blend products and their liking scores were not statistically different. The liking scores in the social setting for both products were not statistically different from liking score of product V (Vitamilk) in the laboratory setting. Both products Q (T37.5%, C25%, P37.5%) and I (T25%, C50%, P25%) contain coconut, tiger nuts and peanuts milk whilst products C (M37.5%, C25%, P37.5 %), N (M37.5%, C25%, P37.5 %) and P(M25%, C50%, P25 %) the least liked products, all contained melon seeds milk. 4.5.2 Product Maps Generalized Procrustes Analysis (GPA) was used to create a two-dimensional consensus map of individual assessor scores for liking and difference from reference. The GPA equalizes the scores and scale usage from different assessors to create a consensus of the scores and uses individual 67 University of Ghana http://ugspace.ug.edu.gh scores to account for variability. Figures 4.5.2 and 4.5.3 show GPA consensus product maps with areas of innovation, that is, the area between the liking and difference axes. Biplot (axes F1 and F2: 100.00 %) 4 3 2 Q Liking V Difference 1 I B 0 D -1 P C -2 BC -3 -4 -3 -2 -1 0 1 2 3 4 5 F1 (88.18 %) Figure 4.5 2 Product map for the social setting Legend: C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q(25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk); V(Vitamilk) and BC (Even UHT full cream milk with added sugar). 68 F2 (11.82 %) University of Ghana http://ugspace.ug.edu.gh Biplot (axes F1 and F2: 100.00 %) 4 3 2 Difference I V Liking 1 Q 0 N P D -1 C -2 BC -3 -4 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 F1 (91.28 %) Figure 4.5 3 Product map for Laboratory setting Legend: C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q(25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk); V(Vitamilk) and BC (Even UHT full cream milk with added sugar). The product map from the social setting explained 39.3% (Rc=0.393) of the variability in the original data whilst the product map from the laboratory explained 53.5% (Rc=0.535) of the variability in the original data. This means that in the laboratory you get better repeatability or reproducibility of the product map. In both the social and laboratory tests, the products were loaded into both maps the same way: for both maps, products Q (T37.5%, C25%, P37.5%) and I (T25%, C50%, P25%), loaded in the innovation area, which is the region between the liking and the difference axes. Although they are considered different from the reference, they are also liked by consumers. They are comparable to the already commercial product V (Vitamilk) which also 69 F2 (8.72 %) University of Ghana http://ugspace.ug.edu.gh loaded in the innovation area and with further optimization could be a potential new product for the plant milk market. The original 3-blend milk of peanut, coconut and melon seed can be said to have been improved upon using consumer insight to replace the melon seed with tiger nut and roasted peanuts as their acceptability was comparable to Vitamilk (the most widely consumed plant milk in Ghana). Consumer acceptability test using BIB provided further insight as to the levels of raw materials that is driving liking. Between products Q (T37.5%, C25%, P37.5%) and I (T25%, C50%, P25%), product Q(T37.5%, C25%, P37.5%) may have more potential as a new plant milk product than product I, because product Q is not as different from the commercial product V (Vitamilk) and BC (the dairy milk product) since plant milk that is similar to dairy milk is more accepted by consumers (Diarra et al, 2005). 4.6 Consumer acceptance test of all 19 formulation prototypes using the BIB design 4.6.1 Mean liking scores from consumer test Table 4.5 shows the mean liking scores for the 3-blend products used in the consumer acceptance test using the BIB design 70 University of Ghana http://ugspace.ug.edu.gh Table 4 5 Mean liking scores for 3-blend products Sensory modalities Product Overall Appearance Flavor Mouthfeel Consistency Aftertaste liking code 4.11±1.83d 6.44±1.24ab 5.00±1.73b 5.11±2.67b 5.78±2.82a 4.11±1.69bc C 6.00±0.50abc 5.77±1.30bc 5.88±1.36ab 5.44±2.13ab 6.00±2.35a 5.78±0.97abc D I 5.33±2.29bcd 6.44±1.33ab 5.33±2.29ab 6.00±1.73ab 5.89±1.76a 6.00±2.35ab 6.56±1.88ab 7.56±1.13a 6.00±2.00ab 5.78±2.90ab 5.78±2.77a 5.11±3.22ab N 4.44±1.67cd 4.78±1.64c 4.44±2.30b 5.11±1.97b 5.00±2.60a 3.78±2.49c P 7.44±1.74a 7.11±1.45a 7.00±1.80a 7.44±1.24a 6.55±2.46a 7.00±2.45a Q C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q (25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk). From Table 4.5, liking scores were generally low for all the modalities measured (overall liking, appearance, flavor, mouth feel, consistency and aftertaste). The highest score across the modalities was the liking score of 7.56±1.13 for the appearance of product N (M25%, C50%, P25%) (this on the 9- point hedonic scale is interpreted as “liked moderately”). Product Q (T37.5%, C25%, P37.5 %,) had the highest overall liking score of 7.44±1.74, but its liking score was not statistically different from the overall liking scores for products D (M25%, C50%, T25%) and N (M25%, C50%, P25%). Product C (M37.5%, C25%, P37.5 %,) had the lowest overall liking score of 4.11±1.83, but its overall liking score was not statistically different from the overall liking scores for products P (M37.5%, C25%, T37.5 %) and I (T25%, C50%, P25 %). Product N (M25%, C50%, P25 %) had the highest liking 71 University of Ghana http://ugspace.ug.edu.gh score for appearance (7.56±1.13), but its liking score was not statistically different from the liking scores for products Q (T37.5%, C25%, P37.5%), I (T25%, C50%, P25 %) and C (M37.5%, C25%, P37.5 %). The products with the lowest liking scores for appearance were products P (M37.5%, C25%, T37.5 %) and D (M25%, C50%, T25 %) but the liking score for product D (M25%, C50%, T25 %) was not statistically different from products C (M37.5%, C25%, P37.5 %) and I (T25%, C50%, P25 %). The product with the highest liking score for flavor was product Q (T37.5%, C25%, P37.5 %) with a flavor liking score of 7.00±1.80, this score was not statistically different from the liking scores of products D (M25%, C50%, T25 %), I (T25%, C50%, P25 %) and N (M25%, C50%, P25 %). The product with the lowest liking score for flavor was product P (M37.5%, C25%, T37.5 %) with a flavor liking score of 4.44±2.30, but this liking score was not statistically different from the liking scores of products D (M25%, C50%, T25 %,), C (M37.5%, C25%, P37.5 %), I (T25%, C50%, P25 %) and N (M25%, C50%, P25 %). For mouth feel, product Q (T37.5%, C25%, P37.5 %) had the highest liking score of 7.44±1.24, but this score was not statistically different from the scores for products D (M25%, C50%, T25 %), I (T25%, C50%, P25 %) and N (M25%, C50%, P25 %). The products with the lowest liking scores for mouth feel were products P (M37.5%, C25%, T37.5 %) and C (M37.5%, C25%, P37.5 %), they had flavor liking scores of 5.11±1.97 and 5.11±2.67 respectively. Their liking scores were not statistically different from the liking scores of products D (M25%, C50%, T25 %), I (T25%, C50%, P25 %) and N (M25%, C50%, P25 %). For consistency, product I (T25%, C50%, P25 %) had the highest liking score of 5.89±1.76, its liking scores were not statistically different from the liking scores of the other products. For aftertaste, the product with the highest liking score was product Q (T37.5%, C25%, P37.5 %), with a liking score of 7.00±2.45. This liking score however was not statistically different from the liking score for products D (M25%, C50%, T25 %), I (T25%, C50%, P25 %) and N (M25%, C50%, P25 %). Product C (M37.5%, C25%, P37.5 %), had the lowest liking score for aftertaste (3.78±2.49), this score was however not 72 University of Ghana http://ugspace.ug.edu.gh statistically different from the liking scores for products D (M25%, C50%, T25 %), and P (M37.5%, C25%, T37.5 %) 4.6.2 Effect of the ingredients on the liking scores for the overall liking, appearance, flavor, mouth feel, consistency and aftertaste Figures 4.6.1 to 4.6.6 show Cox response trace plots that show the effect of the various ingredients on the liking scores for the overall liking, appearance, flavor, mouth feel, consistency and aftertaste. Overall liking Figure 4.6 1 Cox response trace plot for overall liking 73 University of Ghana http://ugspace.ug.edu.gh Figure 4.6.1 shows the effect of the ingredients on overall liking of the products, as coconut milk and peanut milk are increased in the product, it leads to an increase in overall liking scores, whilst tiger nut milk increases the overall liking scores to a point and then decreases it as the proportion of tiger nut milk increases. Melon seeds milk leads to a decrease in the overall liking scores of the product as its proportion increases. Products C (M37.5%, C25%, P37.5 %) and P (M37.5%, C25%, T37.5 %) had the lowest mean liking scores for overall liking, both products had high quantities of melon seeds milk and low amounts of coconut milk, so the low liking scores were expected. Products Q (T37.5%, C25%, P37.5 %) (one of the innovative products according to RPM), N (M25%, C50%, P25 %) and D (M25%, C50%, T25 %) had the highest overall liking scores which were not statistically different. The absence of melon seeds milk in product Q (T37.5%, C25%, P37.5 %), and the high quantities of coconut milk in products N (M25%, C50%, P25 %) and D (M25%, C50%, T25 %) could account for these liking scores. Products C (M37.5%, C25%, P37.5 %) P (M37.5%, C25%, T37.5 %) and I (T25%, C50%, P25 %) had the lowest overall liking scores. The high quantities of melon seeds milk and low amounts of coconut milk in products C and P could have led to these low scores. Even though product I (T25%, C50%, P25 %,) had a high quantity of coconut milk and did not contain melon seeds milk, it had low overall liking scores. This is however an anomaly from the effect of the ingredients on liking scores. 74 University of Ghana http://ugspace.ug.edu.gh Appearance Figure 4.6 2 Cox response trace plot for appearance Figure 4.6.2 shows the effect of the ingredients on appearance liking scores of the products. Melon seeds and tiger nuts milk decrease the liking scores as their proportion increases. Tiger nut milk decreases the liking scores more which could be due to its darker color compared to the rest of the plant milks. Coconut milk increases the liking scores up to a point and reduces it as its quantity increases. However, peanuts milk increases the liking scores as its proportion increases. Products N (M25%, C50%, P25 %), Q (T37.5%, C25%, P37.5 %), I (T25%, C50%, P25 %), and C (M37.5%, C25%, P37.5 %) had the highest liking scores for appearance. The presence of coconut milk and peanut milk in all these formulations could have led to the high liking scores for appearance, any negative effect caused by high amounts of coconut, melon seeds and tiger nuts milk might have been mitigated by the presence of peanuts milk. The products with the lowest liking scores for appearance were products P (M37.5%, C25%, T37.5 %) and D (M25%, C50%, T25 %). The high quantity of coconut milk, coupled with the presence of melon seeds milk and tiger nuts milk in sample 75 University of Ghana http://ugspace.ug.edu.gh product D (M25%, C50%, T25 %) could have led to the low liking scores for its appearance. For product P (M37.5%, C25%, T37.5 %), the high quantities of melon seeds milk and tiger nuts milk could have led to the low appearance liking scores. Flavor Figure 4.6 3 Cox response trace plot for flavor Figure 4.6.3 shows the effect of the ingredients on flavor liking scores of the products. Melon seeds milk decreases the liking scores as its proportion increases, but appears to increase it as its proportion increases to a point, whilst the effect of tiger nut milk is neutral to a point, then decreases the liking scores for flavor as its proportion increases. Coconut milk and peanut milk both increase the liking scores as their proportion increases. The products with the highest liking scores for flavor were Q (P37.5%, C25%, T37.5%), D (M25%, C50%, T25%), I (P25%, C50%, T25%) and N (M25%, C50%, P25%). The presence of peanuts and coconut milk and absence of melon seeds milk in products Q (P37.5%, C25%, T37.5 %) and I (P25%, C50%, T25 %) could have led to the higher flavor liking 76 University of Ghana http://ugspace.ug.edu.gh scores for these products. The high quantities of coconut milk in products D (M25%, C50%, T25%) and N (M25%, C50%, P25%) could have resulted in higher liking scores for flavor eve mitigating the negative effect of melon seeds milk on flavor liking scores. The products with the lowest liking scores for flavor were products C (M37.5%, C25%, P37.5 %) and P (M37.5%, C25%, T37.5%), these products both contained high quantities of melon seeds milk (37.5%) and this could have led to the low liking scores for flavor. Mouth feel Figure 4.6 4 Cox response trace plot for mouth feel Figure 4.6.4 shows the effect of the ingredients on mouth feel liking scores of the products. Melon seeds milk decreases the liking scores as its proportion increases, but appears to increase it as its proportion increases to a point, whilst tiger nut milk increases the scores to a point, then decreases the liking scores as its proportion increases. Coconut milk and peanut milk both increase the liking scores as their proportion increases. The products with the highest liking scores for mouth feel were Q (P37.5%, C25%, T37.5%), D (M25%, C50%, T25%), I (P25%, C50%, T25%) and N 77 University of Ghana http://ugspace.ug.edu.gh (M25%, C50%, P25%). The high quantities of coconut milk in products D (M25%, C50%, T25%), I (P25%, C50%, T25%) and N (M25%, C50%, P25%) and the high quantity of peanut milk in product Q (P37.5%, C25%, T37.5%) could have accounted for the high mouth feel liking scores for these products. The products with the lowest liking score for flavor were products C (M37.5%, C25%, P37.5 %) and P (M37.5%, C25%, T37.5 %). These products both contained high quantities of melon seeds milk (37.5%) which could have led to the low liking scores for mouth feel. Consistency Figure 4.6 5 Cox response trace plot for consistency Figure 4.6.5 shows the effect of the ingredients on consistency liking scores of the products. Melon seeds milk decreases the liking scores as its proportion increases, but appears to increase it as its proportion increases to a point, while the effect of tiger nut milk is neutral to a point, then decreases the liking scores for consistency as its proportion increases. Coconut milk and peanut milk both increase the liking scores as their proportion increases. Even though all the products had consistency liking scores below 7 (like moderately). Their consistency scores were 78 University of Ghana http://ugspace.ug.edu.gh statistically not different from each other. The effect of the various proportions of ingredients on consistency did not result in statistically different liking scores. After taste Figure 4.6 6 Cox response trace plot for aftertaste Figure 4.6.6 shows the effect of the ingredients on the aftertaste liking scores of the products. Melon seeds milk decreases liking scores for aftertaste as its proportion increases, while tiger nuts milk increases the liking scores to a point and starts decreasing it as its proportions increases. Peanut milk and coconut milk on the other hand increase the liking scores. The products with the highest liking scores for after taste were products Q (P37.5%, C25%, T37.5%), D (M25%, C50%, T25%), I (P25%, C50%, T25%) and N (M25%, C50%, P25%).The high quantities of coconut milk in products D (M25%, C50%, T25%), I (P25%, C50%, T25%) and N (M25%, C50%, P25%) and the high quantity of peanut milk in product Q (P37.5%, C25%, T37.5%) could have accounted for the high aftertaste liking scores for these products. The products with the lowest liking scores for after 79 University of Ghana http://ugspace.ug.edu.gh taste were products C (M37.5%, C25%, P37.5 %) and P (M37.5%, C25%, T37.5 %), the high quantities of melon seeds milk could have led to the low liking scores for aftertaste. 4.6.3 Predicting liking scores for optimum formulations Overlaid contour plots generated from regression analysis of consumer acceptance data were used to predict liking scores for optimum formulations. This was done by predicting liking scores in an optimum region when one ingredient was held at zero percent. Figures 4.6.7-4.6.9 show overlaid contour plots generated when peanuts milk, melon seeds milk and tiger nuts milk were held at zero percent. Figure 4.6 7 Overlaid contour plot for formulation when peanuts milk is held at 0% 80 University of Ghana http://ugspace.ug.edu.gh Figure 4.6 8 Overlaid contour plot for formulation when melon seeds milk is held at 0% Figure 4.6 9 Overlaid contour plot when tiger nut milk is held at 0% 81 University of Ghana http://ugspace.ug.edu.gh The white regions in Figures 4.6.7-4.6.9 are the optimum regions; these are where the proportions of the various ingredients will lead to the highest acceptability of the formulation (i.e. a value between 7 and 9 on the 9-point hedonic scale based on the criteria set in the analysis). An ingredient that has an optimum region is driving liking for a particular blend and the ingredient with the biggest optimum region in a blend is the ingredient that is increasing acceptance the most. In Figure 4.6.7, the blend contains melon seeds, coconut and tiger nuts milk, coconut milk shows an optimum region, this means that coconut milk is driving liking for this blend. In Figure 4.6.8, when melon seeds milk is held at zero percent, peanuts milk and coconut milk are showing optimum regions with coconut milk showing a larger optimum region meaning coconut milk is contributing more to liking. Lastly, Figure 4.6.9 shows the plot when tiger nuts milk is held at zero percent. In the blend containing peanuts, coconut and melon seeds milk, peanuts and coconut milk are responsible for liking as they show optimum regions. Tables 4.6 and 4.7 show optimum formulations derived from the optimum regions of the three plots and their predicted liking scores. 82 University of Ghana http://ugspace.ug.edu.gh Table 4 6 Optimum formulations from overlaid contour plots Percentage of each component (%) Formulation Product Coconut Melon seeds Peanuts Tiger nuts code milk milk milk milk 1 C 26.2 0.91 72.90 0 2 D 93.7 1.18 0 5.1 3 I 88.27 0 9.23 2.50 4 N 94.0 2.30 3.66 0 5 P 90.55 2.05 0 7.40 6 Q 71.48 0 26.00 2.50 Table 4 7 Predicted liking scores at optimum regions Sensory Modalities Product Overall Appearance Flavor Mouth feel Consistency Aftertaste liking code C 7.2 7.5 7.0 7.3 7.5 7.0 7.7 7.3 7.7 7.5 7.3 7.6 D I 7.7 7.3 7.3 7.6 7.2 7.4 7.4 7.5 7.4 7.5 7.1 7.2 N P 7.5 7.3 7.3 7.5 7.2 7.3 Q 7.3 7.5 7.1 7.4 7.5 7.2 Legend: C (0.91% melon seeds milk, 26.2% coconut milk, 72.9% peanuts milk); D (1.18% melon seeds milk, 93.7% coconut milk, 5.1% tiger nuts milk); I (88.27% coconut milk, 2.5% tiger nuts milk, 9.23% peanuts milk); N 83 University of Ghana http://ugspace.ug.edu.gh (2.30% melon seeds milk, 94.03% coconut milk, 3.66% peanuts milk); P (2.05% melon seeds milk, 90.55% coconut milk, 7.4% tiger nuts milk) Q(71.48% coconut milk, 2.50% tiger nuts milk, 26% peanuts milk). The amount of coconut milk and peanuts milk are higher in the optimum formulations compared to the initial formulations as these drive acceptance of the blends. Product Q (T37.5%, C25%, P37.5 %) mean liking scores are almost the same as the predicted liking scores. Optimum formulations for products D (M25%, C50%, T25 %) and I (T25%, C50%, P25 %) with high amounts of coconut milk have the highest predicted overall liking scores of 7.7. This could be because of the high quantities of coconut milk in their formulations, 93.70% and 88.27% respectively. Product P (M37.5%, C25%, T37.5%) and N (T25%, C50%, P25%) which also have high amounts of coconut milk in their optimum formulations have the second (7.5) and third highest (7.4) predicted overall liking scores respectively even more than product Q’s predicted overall liking score of (7.3). For appearance liking, optimum formulations for products N (T25%, C50%, P25 %), Q (T37.5%, C25%, P37.5 %) and C (M37.5%, C25%, P37.5 %) have the highest predicted liking scores because they have high amounts of either coconut or peanut milk. Optimum formulations for products C (M37.5%, C25%, P37.5%) contains 72.90% peanut milk, Q (T37.5%, C25%, P37.5 %) contains 71.48% coconut milk and N (T25%, C50%, P25 %) contains 94.0% coconut milk. Their predicted liking scores are 7.5. The optimum formulation for product D (M25%, C50%, T25 %) has the highest predicted flavor liking score of (7.7) because it contains 93.7% coconut milk in its optimum formulation. Product I (T25%, C50%, P25 %) has the highest predicted liking scores for mouth feel (7.6) because it contains a high quantity of coconut milk (88.27%) and 9.31% peanut milk in its optimum formulation. Products C (M37.5%, C25%, P37.5 %) and Q (T37.5%, C25%, P37.5 %) have the highest predicted score (7.5) for consistency because they contain high quantities of peanut milk and coconut milk in their optimum 84 University of Ghana http://ugspace.ug.edu.gh formulations. Product C (M37.5%, C25%, P37.5 %) contains 72.70% peanut milk and 26% coconut milk, whilst product Q (T37.5%, C25%, P37.5 %) contains 71.40% coconut milk and 26% peanuts milk. Product D (M25%, C50%, T25 %) had the highest predicted score for aftertaste (7.6) because it contains a high quantity of coconut milk (93.7%) and also contains peanut milk in its optimum formulation. 4.6.4 Analysis of comments from the Balance Incomplete Block Design Comments analysis can be used to characterize products when the traditional Quantitative Descriptive Analysis (QDA®) cannot be performed due to economic or time constraints (Symoneaux et al., 2012). Figure 4.6.10 shows the output from correspondence analysis carried out on the contingency table produced from the consumer comments for the 3-blend formulations 85 University of Ghana http://ugspace.ug.edu.gh Asymmetric row plot (axes F1 and F2: 54.65 %) 1.5 D L_No_Sugar P D_Not_SweetL_Aroma 1 D_Fante Kenkey D_Color L_Creaminess L_Flavor D_Flavor 0.5 D_Sour L_Consistency D_Taste D_MouthfeelC D_Appearance D_Bitter Aftertaste L_Creamy 0 D_Aftertaste L_Peanut Taste L_Sweet L_Taste L_Appearance D_Bitter D_CoarsLe_ Mouthfeel -0.5 I PartiLc_leAsftertaste Q D_ALr_oNmoat_Bitter L_Color L_Tiger nut … -1 D_Too_Sweet NL_Smooth L_Coconut Taste -1.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 F1 (31.10 %) Rows Columns Figure 4.6 10 Correspondence Analysis for 3- blend formulations Legend: D – Dislike comment; L – Like comments. C (37.5% melon seeds milk, 25% coconut milk, 37.5% peanuts milk); D (25% melon seeds milk, 50% coconut milk, 25% tiger nuts milk); I (50% coconut milk, 25% tiger nuts milk, 25% peanuts milk); N (25% melon seeds milk, 50% coconut milk, 25% peanuts milk); P (37.5% melon seeds milk, 25% coconut milk, 37.5% tiger nuts milk) Q(25% coconut milk, 37.5% tiger nuts milk, 37.5% peanuts milk) For the three blend beverages that were used for the RPM, the output for the correspondence analysis grouped the innovative products, I (T25%, C50%, P25%) and Q (T37.5%, C25%, P37.5%) in the same quadrant, products P(M37.5%, C25%, T37.5%) and D(M25%, C50%, T25%) loaded in the same quadrant whilst products N(M25%, C50%, P25%) and C(M37.5%, C25%, P37.5%) were not grouped with any of the other products and were placed singly. The like terms for products I (T25%, C50%, P25 %) and Q (T37.5%, C25%, P37.5%) were tiger nut taste, not bitter, mouth feel, aftertaste, mouth 86 F2 (23.55 %) University of Ghana http://ugspace.ug.edu.gh feel, sweet and taste. The dislike terms used for them were coarse particles, aroma and too sweet. The taste and aftertaste of these products were liked because they did not contain melon seeds milk which decreases liking for flavor and aftertaste. The like terms for products P (M37.5%, C25%, T37.5 %) and D (M25%, C50%, T25 %) were aroma, flavor and no sugar. The dislike terms used for them included not sweet, Fante kenkey, color, flavor, sour, taste and appearance. Product D (M25%, C50%, T25 %) contains 50% coconut and this could have increased the liking for flavor as some consumers liked the flavor of this product. Some consumers disliked the flavor of both products P (M37.5%, C25%, T37.5 %) and D (M25%, C50%, T25%) this was not expected for product D as it contained coconut milk but the high amount of melon seeds milk in product P could account for the dislike for the flavor. The appearance and color of both products were not liked. Product P (M37.5%, C25%, T37.5%) had the lowest L value amongst the 3-blend beverages of 75.91±0.13, but product D (M25%, C50%, T25%) had an L value of 80.06±0.77 as it contained less tiger nuts milk compared to product P (M37.5%, C25%, T37.5%). Their color and appearance were not liked because of the dark color imparted by tiger nuts milk. The like terms for product C (M37.5%, C25%, P37.5%) were creamy, peanut taste, consistency and creaminess and the dislike terms were aftertaste, bitter and mouth feel. The like terms for product N (M25%, C50%, P25 %) were smooth, coconut taste, appearance and color. The only dislike term was bitter. The high amounts of melon seeds milk in both products could have accounted for both products having dislike terms bitter and aftertaste. Product C (M37.5%, C25%, P37.5 %,) had a high amount of peanut milk which accounted for the like term peanut taste, whilst the high quantity of coconut milk in product N (M25%, C50%, P25 %) resulted in the like term coconut taste, both peanut and coconut milk increase liking for flavor. Coconut milk and peanut milk both increase liking for mouth feel, which accounts for the like term smooth for product N (M25%, C50%, P25%). Product N (M25%, C50%, P25 %), 87 University of Ghana http://ugspace.ug.edu.gh had the highest L value of 85.54±0.50 amongst the 3-blend beverages which could account for why the color of this product is liked. 5.0 CONCLUSIONS Consumer input through focus group discussions provided useful insight to guide the optimization of the initial plant milk beverages which led to products that were accepted by consumers: peanuts were roasted before processing and tiger nut milk was added to the formulations, the minimum quantity of coconut was kept constant at 25%. Two of the optimized products (containing peanuts, tiger nuts and coconut milk) were the most liked amongst the optimized products and were perceived as innovative products by consumers, their liking scores were comparable to Vitamilk, a commercially successful plant milk beverage. Characterization of the optimized 3-blend beverages by RPM provided a quick way to visualize which of the optimized products was accepted by consumers, was innovative and at the same time provided information about how the products compared with an already accepted plant milk product (Vitamilk). The consumer acceptability test using BIB was successfully used to pin-point which of the components in the 3-blend beverages were responsible for acceptability of the innovative 3-blend beverages and the results from the BIB agreed with the results from the RPM, as both pointed to the conclusion that coconut milk and peanut milk were responsible for the acceptability of the blends. Combining three plant varieties could lead to products with proximate composition comparable to dairy milk. 88 University of Ghana http://ugspace.ug.edu.gh 6.0 RECOMMENDATIONS • The optimum formulations obtained using BIB could be processed and used for a consumer acceptability test to confirm their degree of acceptance by consumers • A quantitative descriptive test could be done to fully understand the sensory properties of the 3-blends that drive liking • Further studies should be done to ascertain the shelf-life and suspension stability of the plant milk beverages. 89 University of Ghana http://ugspace.ug.edu.gh 7.0 REFERENCES Abiodun, O. A and Adeleke, R.O. 2010. Comparative Studies on Nutritional Composition of Four Melon Seeds Varieties. Pakistan Journal of Nutrition. 9(9), 905-908. http://docsdrive.com/pdfs/ansinet/pjn/2010/905-908.pdf Aboulfazli, F., Baba A.S. and Misran, M. (2014). Effect of Vegetable Milks on the Physical and Rheological Properties of Ice Cream. Food Science and Technology Research, 20 (5), 987-996. Available:file:///C:/Users/User%201/Downloads/EffectofVegetableMilksonthePhysicalandRheol ogicalPropertiesofIceCream.pdf Aboulfazli, F., Shori, A.B and Baba, A.S. (2016). Effects of the replacement of cow milk with vegetable milk on probiotics and nutritional profile of fermented ice cream. LWT - Food Science and Technology, 70, 261-270. http://dx.doi.org/10.1016/j.lwt.2016.02.056. Adjepong, Mary., Valentini, K., Pickens, C.A., Li W., Appaw, W. and Fenton, J. (2017). Quantification of fatty acid and mineral levels of selected seeds, nuts, and oils in Ghana. Journal of Food Composition and Analysis, 59, 43-49. http://dx.doi.org/10.1016/j.jfca.2017.02.007. Aguilar, N., Albanell, E., Miñarro, B., and Capellas, M. (2015). Chickpea and tiger nut flours as alternatives to emulsifier and shortening in gluten-free bread. LWT - Food Science and Technology, 62(1), 225-232. http://dx.doi.org/10.1016/j.lwt.2014.12.045. Aidoo, H., Sakyi-Dawson, E., Tano-Debrah. K., Saalia, K. F. (2010). Development and characterization of dehydrated peanut–cowpea milk powder for use as a dairy milk substitute in chocolate manufacture. Food Research International, 43(1), 79-85. http://dx.doi.org/10.1016/j.foodres.2009.08.018. Aidoo, H., Sakyi-Dawson, E., Abbey, L., Tano-Debrah K., Saalia K.F. (2012). Optimization of chocolate formulation using dehydrated peanut–cowpea milk to replace dairy milk. Journal of the Science of Food and Agriculture, 92(2), 224-231. https://doi.org/10.1002/jsfa.4563 Akubor P. I (1998) Physicochemical and sensory characteristics of melon seed milk. J Food Sci Technol-Mysore 35(1): 93-97. Akubor, P. I. and Ogbadu, R. L. (2003). Effects of processing methods on the quality and acceptability of melon milk. Plant Foods for Human Nutrition, 58 (1), 1–6, https://doi.org/10.1023/A:1024063105507. 90 University of Ghana http://ugspace.ug.edu.gh Albors, A., Raigon, M.D., García-Martinez, M.D., Martín-Esparza, M. E. (2016). Assessment of techno-functional and sensory attributes of tiger nut fresh egg tagliatelle. LWT - Food Science and Technology, 74, 183-190. http://dx.doi.org/10.1016/j.lwt.2016.07.047. Aluko R.E. (2018) Food protein-derived peptides: Production, isolation, and purification In Woodhead Publishing Series in Food Science, Technology and Nutrition. Proteins in Food Processing (Second Edition). 389-412. https://doi.org/10.1016/B978-0-08-100722-8.00016-4. Antúnez, L., Vidal, L., Saldamando, L., Giménez, A., Ares, G. (2017).Comparison of consumer- based methodologies for sensory characterization: Case study with four sample sets of powdered drinks. Food Quality and Preference, 56(A), 149-163. https://doi.org/10.1016/j.foodqual.2016.09.013. AOAC - Association of Official Analytical Chemists. (2005). Official methods of analysis of AOAC international (18th ed.). Arlington: AOAC. Ares. G., Giménez, A., Barreiro, C. and Gámbaro, A. (2010a). Use of an open-ended question to identify drivers of liking of milk desserts.Comparison with preference mapping techniques. Food Quality and Preference, 21(3), 286-294, https://doi.org/10.1016/j.foodqual.2009.05.006. Ares. G., Deliza R., Barreiro, C., Giménez, A and Gámbaro A. (2010b). Comparison of two sensory profiling techniques based on consumer perception. Food Quality and Preference, 21 (4), 417-426, https://doi.org/10.1016/j.foodqual.2009.10.006. Ares, G., Varela, P., Rado, G. and Giménez, A. (2011). Are consumer profiling techniques equivalent for some product categories? The case of orange‐flavoured powdered drinks. International Journal of Food Science & Technology, 46: 1600-1608. Doi:10.1111/j.1365- 2621.2011.02657.x Ares, G., Antúnez, L., Oliveira, D., Alcaire, F., Giménez A., Berget, I., Næs, T. and Varela, P. (2015). Pole selection in Polarized Sensory Positioning: Insights from the cognitive aspects behind the task, Food Quality and Preference, 46, 48-57. https://doi.org/10.1016/j.foodqual.2015.07.003. 91 University of Ghana http://ugspace.ug.edu.gh Arya, S.S., Salve, A.R. and Chauhan, S. (2016). Peanuts as functional food: a review. Journal of Food Science and Technology. 53(1), 31-41. https://doi.org/10.1007/s13197-015-2007-9 . Ayeh-Kumi, P. F., Tetteh-Quarcoo, P. B., Duedu, K. O., Obeng, A. S., Addo-Osafo, K., Mortu, S. and Asmah, R. H. (2014). A survey of pathogens associated with Cyperus esculentus L (tiger nuts) tubers sold in a Ghanaian city. BMC Research Notes, 7, 343. http://doi.org/10.1186/1756- 0500-7-343. Bankole, S.A., Osho, A., Joda A.O. and Enikumenim, O.A. (2005).The Effect of drying method on the quality and storability of ‘egusi’ melon seeds (Colocynthis citrullus L). African Journal of Biotechnology, 4(8) 799-803. Available: https://www.ajol.info/index.php/ajb/article/view/15185/59502. Bastıoğlu Z. A., Tomruk, D., Koç, M. and Ertekin, F. K. (2016). Spray dried melon seed milk powder: physical, rheological and sensory properties. Journal of Food Science and Technology, 53(5), 2396–2404. http://doi.org/10.1007/s13197-016-2214-z. Belewu M. A and Belewu K.Y (2007). Comparative physico-chemical evaluation of tiger nut, soybean and coconut milk sources. International Journal of Agriculture and Biology, 9(5), 785- 787. Blay, M .Y. (2012). Consumer preference mapping using wine show derived data. PhD Thesis. Charles Sturt University, Australia. Boutrolle, I., Arranz, D., Rogeaux, M. and Delarue, J. (2005). Comparing central location test and home use test results: Application of a new criterion. Food Quality and Preference, 16 (8), 704-713. https://doi.org/10.1016/j.foodqual.2005.03.015. Bruzzone, F., Vidal, L., Antúnez, L., Giménez, Ana., Deliza, R. and Ares, G. (2015). Comparison of intensity scales and CATA questions in new product development: Sensory characterisation and directions for product reformulation of milk desserts, Food Quality and Preference, 44, 183-193. https://doi.org/10.1016/j.foodqual.2015.04.017. Cadena, R.S., Caimi, D., Jaunarena, I., Lorenzo, I., Vidal, L., Ares, G., Deliza, R. and Giménez, A. (2014). Comparison of rapid sensory characterization methodologies for the development of functional yogurts. Food Research International, 64, 446-455. https://doi.org/10.1016/j.foodres.2014.07.027. Codina-Torrella, I., Guamis, B., and Trujillo, A. J. (2015). Characterization and comparison of tiger nuts (L.) from different geographical origin. Industrial Crops and Products, 65, 406-414. http://dx.doi.org/10.1016/j.indcrop.2014.11.007. 92 University of Ghana http://ugspace.ug.edu.gh Costa, A.I.A. and Jongen, W.M.F. (2006). New insights into consumer-led food product development. Trends in Food Science & Technology, 17(8), 457-465. https://doi.org/10.1016/j.tifs.2006.02.003. De Graaf, C., Cardello, A. V., Kramer, F. M., Lesher, L. L., Meiselman H. L. and Schutz H.G.(2005). A comparison between liking ratings obtained under laboratory and field conditions: the role of choice. Appetite, 44 (1), 15-22. https://doi.org/10.1016/j.appet.2003.06.002 Dhakal, S., Liu, C., Zhang, Y., Roux, K. H., Sathe, S. K. and Balasubramaniam, V. M. (2014) Effect of high pressure processing on the immunoreactivity of almond milk. Food Research International, 62, 215-222. https://doi.org/10.1016/j.foodres.2014.02.021. Diarra, K., Nong, Z. G. and Jie, C. (2005). Peanut milk and peanut milk-based products production: A review. Critical Reviews in Food Science and Nutrition, 45, 405–423. https://doi.org/10.1080/10408390590967685 Dooley, L., Lee,Y. and Meullenet, J. (2010). The application of check-all-that-apply (CATA) consumer profiling to preference mapping of vanilla ice cream and its comparison to classical external preference mapping. Food Quality and Preference, 21(4), 394-401 https://doi.org/10.1016/j.foodqual.2009.10.002. Ekpong, A. Ngarmsak, T. and Winger, R. J. (2006). Comparing sensory methods for the optimization of mango gel snacks. Food Quality and Preference, 17(8), 622-628. https://doi.org/10.1016/j.foodqual.2006.06.001. Ezeh, O., Gordon, M.H. and Niranjan, K. (2014). Tiger nut oil (Cyperus esculentus L.): A review of its composition and physico-chemical properties. European Journal of Lipid Science and Technology, 116(7), 783-794. https://doi.org/10.1002/ejlt.201300446. Ezuruike, U. F. and Prieto, J. M. (2014). The use of plants in the traditional management of diabetes in Nigeria: Pharmacological and toxicological considerations. Journal of Ethnopharmacology, 155(2), 857-924. http://dx.doi.org/10.1016/j.jep.2014.05.055. Ganguly, S. (2013). Health benefits of coconut in the Asian cuisines: A review. Journal of Biological and Chemical Research, 30(2), 517-21. Available: http://jbcr.co.in/Current_Issue/Volume%2030%20(2)%20July%20to%20December%20A/Health -Benefits-of-Coconut.pdf. 93 University of Ghana http://ugspace.ug.edu.gh Hamdy, S.M., Shabaan, A.M., Latif, A. K. M. A., Abdel-Aziz A.M. and Amin A.M. (2017). Protective effect of Hesperidin and Tiger nut against Acrylamide toxicity in female rats. Experimental and Toxicologic Pathology, 69(8), 580-588. http://dx.doi.org/10.1016/j.etp.2017.05.004. Hein, K.A. Jaeger, S.R., Carr, B.T. and Delahunty, C.M. (2008). Comparison of five common acceptance and preference methods. Food Quality and Preference, 19(7), 651-661. https://doi.org/10.1016/j.foodqual.2008.06.001. Humiski, L. and Aluko, R. (2007). Physicochemical and Bitterness Properties of Enzymatic Pea Protein Hydrolysates. Journal of Food Science, 72(8), 605-611. https://doi.org/10.1111/j.1750- 3841.2007.00475.x Isanga, J. and Zhang, G. (2009). Production and evaluation of some physicochemical parameters of peanut milk yoghurt. LWT - Food Science and Technology, 42(6), 1132-1138. https://doi.org/10.1016/j.lwt.2009.01.014. Jain, P., Yadav, D. N., Rajput, H., and Bhatt, D. K. (2013). Effect of pressure blanching on sensory and proximate composition of peanut milk. Journal of Food Science and Technology, 50(3), 605–608. http://doi.org/10.1007/s13197-011-0373-5. Jeske, S., Zannini, E. and Arendt, E. K. (2018). Past, present and future: The strength of plant- based dairy substitutes based on gluten-free raw materials. Food Research International, 110, 42-51. https://doi.org/10.1016/j.foodres.2017.03.045. Jeske, S., Zannini, E., & Arendt, E. K. (2016). Evaluation of Physicochemical and Glycaemic Properties of Commercial Plant-Based Milk Substitutes. Plant foods for human nutrition (Dordrecht, Netherlands), 72(1), 26-33. Kane, N., Ahmedna, M. and Yu, J. (2010). Development of a fortified peanut-based infant formula for recovery of severely malnourished children. International Journal of Food Science and Technology, 45(10), 1965-1972. https://doi.org/10.1111/j.1365-2621.2010.02330.x . Kizzie-Hayford, N., Jaros, D., Zahn, S. and Rohm, H. (2016). Effects of protein enrichment on the microbiological, physicochemical and sensory properties of fermented tiger nut milk. LWT - Food Science and Technology, 74, 319-324. https://doi.org/10.1016/j.lwt.2016.07.067. Lasekan, O. and Abdulkarim, S. M. (2012). Extraction of oil from tiger nut (Cyperus esculentus L.) with supercritical carbon dioxide (SC-CO). LWT - Food Science and Technology, 47(2), 287- 292. http://dx.doi.org/10.1016/j.lwt.2012.01.021. 94 University of Ghana http://ugspace.ug.edu.gh Lawal, S.A., Choudhury I.A. and Nukman, Y. (2012). An assessment of the physico-chemical properties of melon seed (Citrullus lanatus) oil as base material for oil-in-water emulsion cutting fluid. Advanced Materials Research, 576, 293-295. https://doi.org/10.4028/www.scientific.net/AMR.576.293. Linnemann, A. R.., Benner, M., Verkerk, R., Martinus, A.J.S. and van Boekel. (2006). Consumer-driven food product development. Trends in Food Science & Technology, 17(4), 184-190. https://doi.org/10.1016/j.tifs.2005.11.015. Mäkinen, O.E., Wanhalinna, V., Zannini, E. and Arendt, E. K. (2016). Foods for special dietary needs: Non-dairy plant-based milk substitutes and fermented dairy type products. Critical Reviews in Food Science and Nutrition, 56(3), 339–349. http://dx.doi.org/10.1080/10408398.2012.761950. Mäkinen, O. E., Uniacke-Lowe, T., O’Mahony J. A. and Arendt, E. K. (2015) Physicochemical and acid gelation properties of commercial UHT-treated plant-based milk substitutes and lactose free bovine milk, Food Chemistry, 168, 630-638. https://doi.org/10.1016/j.foodchem.2014.07.036. Martinez-González, M.A., and Bes-Rastrollo, M. (2011). Nut consumption, weight gain and obesity: Epidemiological evidence, Nutrition, Metabolism and Cardiovascular Diseases, 21, S40-S45, http://dx.doi.org/10.1016/j.numecd.2010.11.005. Meilgaard, M., Civille, G. V. and Carr, B. T. (2006). Sensory evaluation techniques, CRC press. Meiselman, H.L., (2013). The future in sensory/consumer research: ………......evolving to a better science. Food Quality and Preference, 27(2), 208-214, https://doi.org/10.1016/j.foodqual.2012.03.002. Mridula, D., and Sharma, M. (2015). Development of non-dairy probiotic drink utilizing sprouted cereals, legume and soymilk. LWT - Food Science and Technology, 62(1), 482- 487.http://dx.doi.org/10.1016/j.lwt.2014.07.011. Nunes, C. A., Bastos, S. C., Pinheiro, A. C., Pimenta, C. J. And Pimenta, M. E. (2012). Relating Consumer Acceptance to Descriptive Attributes By Three‐Way External Preference Mapping Obtained By Parallel Factor Analysis (Parafac). Journal of Sensory Studies, 27, 209-216. https://doi.org/10.1111/j.1745-459X.2012.00387.x 95 University of Ghana http://ugspace.ug.edu.gh Odoom, E. A (2018). Formulation and evaluation of the physiochemical and sensory properties of vegetable milk from peanuts, melon seeds and coconut. Mphil Thesis. University of Ghana, Ghana. Ogunlade, I, Adeyemi, B. A and Aluko, O.G. (2015). Chemical compositions, antioxidant capacity of Tiger nut (Cyperus esculentus) and potential health benefits. European Scientific Journal 1–8. Available: https://pdfs.semanticscholar.org/287b/36db7c5e83e55dd625385e82e2c2876facad.pdf. Ojieh, G.C., Oluba, O.M., Ogunlowo, Y.R., Adebisi, K.E., Eidangbe, G.O. and Orole, R.T. (2008). Compositional studies of Citrullus lanatus (Egusi melon) seed. The Internet Journal of Nutrition and Wellness, 6(1), 1937-8297.Available: file:///C:/Users/User%201/Downloads/ispub- 67211.pdf Oluba, O.M., Adeyemi, O., Ojieh, G.C. and Isiosio, I. O. (2008) Fatty acid composition of Citrullus lanatus (Egusi melon) oil and its effect on serum lipids and some serum enzymes. The Internet Journal of Cardiovascular Research, 5(2), 1540-2592. Available: https://www.researchgate.net/publication/285321869. Omoni, A. O. and Aluko, R. E. (2005), Soybean Foods and Their Benefits: Potential Mechanisms of Action. Nutrition Reviews, 63, 272-283. https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1753-4887.2005.tb00141.x Onning, G., Akesson, B., €Oste, R. and Lundquist, I. (1998). Effects of consumption of oat milk, soya milk, or cow’s milk on plasma lipids and antioxidative capacity in healthy subjects. Annals of Nutrition and Metabolism. 42, 211–220. Available: https://www.karger.com/Article/Abstract/12736. Onyeike E. N and Acheru G. N. (2002). Chemical composition of selected Nigerian oil seeds and physicochemical properties of the oil extracts, Food Chemistry, 77(4), 431-437, https://doi.org/10.1016/S0308-8146(01)00377-6. Otoo, G. (2017, July). Overview of Dairy Nutrition in Ghana. Paper presented at workshop to develop a conceptual framework on Formative and Feasibility Assessment of Opportunities for Strengthening Dairy Value Chains in Ghana for Improved Maternal and Child Nutrition Outcomes. Accra, Ghana 96 University of Ghana http://ugspace.ug.edu.gh Owusu Yeboah, S., Mitei, Y.C., Ngila, J.C., Wessjohann, L. and Schmidt, J. (2012). Compositional and structural studies of the oils from two edible seeds: Tiger nut, Cyperus esculentum, and asiato, Pachira insignis, from Ghana, Food Research International, 47(2), 259- 266. http://dx.doi.org/10.1016/j.foodres.2011.06.036. Palacios, O. M., Badran, J., Drake, M. A., Reisner, M. and Moskowitz, H. R. (2009). Consumer acceptance of cow’s milk versus soy beverages: Impact of ethnicity, lactose tolerance and sensory preference segmentation. Journal of Sensory Studies. 24, 731–748. Available: http://onlinelibrary.wiley.com/doi/10.1111/j.1745-459X.2009.00236.x/full. Palacios, O. M., Badran, J., Spence, L., Drake, M. A., Reisner, M. and Moskowitz, H. R. (2010). Measuring acceptance of milk and milk substitutes among younger and older children. Journal of Food Science, 75, S522–S526. Available: http://onlinelibrary.wiley.com/doi/10.1111/j.17503841.2010.01839.x/full. Pelsmaeker, S.D., Gellynck, X., Delbaere C., Declercq, N. and Dewettinck, K. (2015). Consumer-driven product development and improvement combined with sensory analysis: A case-study for European filled chocolates, Food Quality and Preference, 41, 20-29. https://doi.org/10.1016/j.foodqual.2014.10.009. Pham, L. J. (2016). Industrial Oil Crops: Chapter 9 - Coconut (Cocos nucifera). https://doi.org/10.1016/B978-1-893997-98-1.00009-9. Pineli, L.L.O., Botelho, R.B.A., Zandonadi R. P., Solorzano J. L., Oliveira G. T., Reis C. E. G. and Teixeira D.S. (2015). Low glycemic index and increased protein content in a novel quinoa milk, LWT - Food Science and Technology, 63 (2), 1261-1267. https://doi.org/10.1016/j.lwt.2015.03.094. Raghavendra, S.N. and Raghavarao, K.S.M.S. (2010). Effect of different treatments for the destabilization of coconut milk emulsion. Journal of Food Engineering, 97(3), 341-347. https://doi.org/10.1016/j.jfoodeng.2009.10.027. 97 University of Ghana http://ugspace.ug.edu.gh Saalia, F.K., Mankanta, C., Budu, A., and Essilfie, G. (2013). Knowledge and consumption patterns of coffee creamers in Accra, Ghana. Nutrition & Food Science, 43(1) 23-30. https://doi.org/10.1108/00346651311295879. Sánchez-Zapata, E., Fernández-López, J. and Pérez-Alvarez, A. J. (2012). Tiger Nut (Cyperus esculentus) Commercialization: Health Aspects, Composition, Properties, and Food Applications. Comprehensive Reviews in Food Science and Food Safety, 11, 366–377. Available: http://onlinelibrary.wiley.com/doi/10.1111/j.1541-4337.2012.00190.x/full. Schifferstein, H.N.J. (2015). Employing consumer research for creating new and engaging food experiences in a changing world. Current Opinion in Food Science, 3 27-32. https://doi.org/10.1016/j.cofs.2014.11.004. Sebastià, N., Meca, G., Soriano, J. M. and Mañes, J. (2012). Presence of emerging mycotoxins in tiger-nuts commercialized in Spain. Food Control, 25(2), 631-635, http://dx.doi.org/10.1016/j.foodcont.2011.11.040. Swati, S., Tyagi, S. K. and Anurag, R. K. (2016). Plant-Based Milk Alternatives an Emerging Segment of Functional Beverages: A Review. Journal of Food Science and Technology. 53(9), 3408–3423. Available: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5069255/ Symoneaux, R., Galmarini, M. V. and Mehinagic, E. (2012). Comment analysis of consumer’s likes and dislikes as an alternative tool to preference mapping: A case study on apples. Food Quality and Preference, 24(1), 59-66. https://doi.org/10.1016/j.foodqual.2011.08.013. Tangsuphoom, N. and Coupland, J. N. (2005), Effect of Heating and Homogenization on the Stability of Coconut Milk Emulsions. Journal of Food Science, 70: e466–e470. Available: http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2621.2005.tb11516.x/full 98 University of Ghana http://ugspace.ug.edu.gh Tansakul, A. and Chaisawang, P. (2006). Thermophysical properties of coconut milk. Journal of Food Engineering, 73(3), 276-280. https://doi.org/10.1016/j.jfoodeng.2005.01.035. Thompson, J., Drake, M., Lopetcharat, K. and Yates, M. (2004), Preference Mapping of Commercial Chocolate Milks. Journal of Food Science, 69: S406-S413. doi:10.1111/j.1365- 2621.2004.tb09958.x Tuso, P.J., Ismail, M.H., Ha, B. P. and Bartolotto, C. (2013). Nutritional Update for Physicians: Plant-Based Diets. The Permanente Journal, 17 (2), 61-66. http://doi.org/10.7812/TPP/12-085. Uruakpa, F.O. and Aluko, R.E (2004). Heat-induced gelation of whole egusi (Colocynthis citrullus L.) seeds (2004). Food chemistry. 87 (3), 349-354. https://doi.org/10.1016/j.foodchem.2003.12.005. Valentin, D., Chollet, S., Lelièvre, M. and Abdi, H. (2012), Quick and dirty but still pretty good: a review of new descriptive methods in food science. International Journal of Food Science & Technology, 47: 1563-1578. doi:10.1111/j.1365-2621.2012.03022.x van Kleef, E., van Trijp, H.C.M. and Luning, P. (2005). Consumer research in the early stages of new product development: a critical review of methods and techniques, Food Quality and Preference, 16(3), 181-201https://doi.org/10.1016/j.foodqual.2004.05.012 99 University of Ghana http://ugspace.ug.edu.gh APPENDIX 1 Recruitment questionnaire used for Focus Group Discussion Are you willing to participate in a 90-minute focus group discussion on plant milk? a. Yes b. No If No, do not continue the discussion If Yes, invite to focus group Q1. Gender (NOTE GENDER, DON’T READ) a. Male b. Female If a in Q1, Check Quota (Code 1) If b in Q1, check Quota (Code 2) Q2. Age a. 18-30 [check code 3] b. 31-49 [check code 4] c. 50+[check code 5] d. Age unknown or do not wish to disclose [ discontinue] Q3. How often did you consume any plant milk in the immediate past month a. Everyday b. Once c. Twice d. Several times e. Never If e, in Q3-Terminate If a or b, c or d continue to Q4 Q4. On average how many liters of plant milk do you consume in a week (note: A bottle of Vitamilk is 300ml and a pack of UHT Soymilk or Alpro is 1000ml) …………………………… Q5. On average how much do you spend on plant milk in a week (in GHS) ................................. Q6. On a scale of 1-7, (where 1 is “Not at all” and 7 is “very much”, how enthusiastic are you about plant milk? (Please circle only 1 option). a. 1-Not at all enthusiastic b. 2-slightly not enthusiastic 100 University of Ghana http://ugspace.ug.edu.gh c. 3-moderately not enthusiastic d. 4-neither un-enthusiastic nor enthusiastic e. 5.slightly enthusiastic f. 6-Moderately enthusiastic g. 7-Very much enthusiastic Q7. Are you allergic/intolerant to any of these ingredients that restrict you from evaluating plant milk? Check all the apply a. Dairy milk b. Melon seeds(Egusi) c. Coconut d. Peanut e. Almonds f. Soybeans g. Other, please specify………………………………………………… If allergic/intolerant to any of these, terminate interview Q8. What is your highest educational level? a. School leaver b. Junior High School c. Senior High School d. Tertiary e. Postgraduate f. Other, please specify…………………………. Q9. How well are you able to communicate (spoken and written) in English language? a. Not at all b. Beginner c. Intermediate d. Advance Terminate if a or b. Continue if c or d You are invited to participate in a focus group discussion that will involve you coming to the Sensory Laboratory of the Department of Nutrition and Food science in the University of Ghana. Participants in the focus group discussion will discuss in a relaxed and informal setting, where every view will be respected. We have ethical clearance from the ECBAS and are mandated to protect our participants’ rights. Codes for quotas required Code 9= 18-30 years mixed group n=10 Code 10= 31-49 years mixed group n=10 Code 11=50+years mixed group n=10 101 University of Ghana http://ugspace.ug.edu.gh Q11. Please indicate if you will be available on the date for your age group a. Friday, 10th February, 2017, Sensory Laboratory, NFS, UG (18-30 years mixed) (2.00pm- 3.30pm) b. Monday, 13th February, 2017, Sensory Laboratory, NFS, UG (31-49 years mixed) (10.00am-11.30am) c. Friday, 17th February, 2017, Sensory Laboratory, NFS, UG (50+ years mixed) (2.00pm- 3.30pm)) If a in Q11, Continue, check Quotas (Code 9) If b in Q11, Continue, check Quotas (Code 9) If c in Q11, Continue, check Quotas (Code 9) PART 2: QUESTIONAIRE Q12. What is your occupation? a. Professional/Managerial/Administrator b. Arts/Communication c. Tradesman/Artisan d. Laborer e. Retired f. Stay at Home g. Between work h. Business owner/Self employed i. Producer (Farmer and Fisherman) j. Uniformed (Military/Policeman/Fire service/Prisons) k. Student l. Other, please specify…………………………………. Q13.What is your ethnicity? a. Ewe b. Akan-Forest areas (Akuapem, Akyem, Ashanti’s, Sefwi, etc.) c. Akan-Coastal areas (Fantes, Nzema, Ahanta etc.) d. Guan e. Northern f. Ga/Adangbe g. Other, please specify…………………………………… Q14a.How many people are in your family? (not including you) a. Two 102 University of Ghana http://ugspace.ug.edu.gh b. Three c. Four d. Five e. Six f. Other, Please specify………………………….. Q14b. How many are Children? a. One b. Two c. Three d. Four e. Five f. Other, Please specify………………………….. Thank you very much for your patience in answering these questions. Based on the information you have provided you qualify /do not qualify to participate in the focus group discussion as described above. You will be contacted a day prior to the test day to remind you of the test. 103 University of Ghana http://ugspace.ug.edu.gh APPENDIX 2 Moderators Guide for focus group discussion MODERATORS GUIDE FOR PLANT MILK FOCUS GROUP INTRODUCTION (2 minutes) My name is Dr. Maame Yaakwaah Adjei and I’m the moderator today. The purpose of this discussion is to talk about Plant Milk. I’ll be asking your opinions and your experiences. All participant comments and opinions will be kept anonymous and confidential. You will be made to sign a consent form before the discussions begin. I am a sensory scientist, who uses sensory and consumer testing to help in new product development and improve existing products. I believe in integrating consumer needs and wants in developing new products. INTRODUCTIONS Assistant moderators GROUND RULES (5 minutes) 1. This session will last about 75-90minutes. 2. This session is being audio taped, I will take this opportunity to ask if there is anyone who objects to being recorded (Explain why?) 3. Observers who are part of the team are present to observe and also take notes. 4. There are no wrong answers in consumer research; we are looking for different points of view. I want to know what your opinions are. 5. Everyone needs to talk but each person doesn’t have to answer each question. 6. Please talk one at a time and in a clear voice, avoid side conversations. It is distracting to the group and I don’t want to miss any of your comments. 7. Exchange points of view with each other – you don’t need to address all answers to me. 8. This is a confidential discussion in that I will not report your names or who said what to your colleagues or supervisors. Names of participants will not even be included in the final report about this meeting. It also means, except for the report that will be written, what is said in this room stays in this room. 104 University of Ghana http://ugspace.ug.edu.gh 9. We stress confidentiality because we want an open discussion. We want all of you to feel free to comment on each other’s remarks without fear your comments will be repeated later and possibly taken out of context. 10. Let me know if you need a break. The bathrooms are [location]. Feel free to enjoy a beverage and a snack. 11. Please put all phones on silent 12. Does anyone have any questions before we begin? BACKGROUND (10 minutes) 1. Please take the first few minutes to get to know the person seated next to you. 2. I’m going to ask you to introduce him/her to the group. I’d like you to tell us something about that person such as their interests and hobbies, if they have a family, and how long they have lived in the area. 3. Introduce your friend by adding an adjective with the letter of their first names (18- 30) QUESTIONS PART 1 (20 minutes) 1. What are the types of plant milk you know of? 2. What are they made of? 3. What plant milks have you drank? 4. What were they made of? 5. Apart from drinking them, what can they be used for? 6. Which people do you think drink plant milk? 7. Why do you like plant milk? What makes you buy plant milk? 8. Are there any perceived health benefits? Activity 1 (10 minutes) You are presented with products of plant milk, what do you like about the appearance (Colour), Aroma, Flavor (including intensity), Texture (thickness or lightness), and Aftertaste. 105 University of Ghana http://ugspace.ug.edu.gh QUESTIONS PART 2 (20 minutes) 9. What did you like about the plant milk you drank? 10. What did you not like about each of them? 11. What sensory qualities do you prefer in plant milk and why? 12. What packaging do you prefer for plant milk? 13. What local ingredients can be used for plant milk? 14. What do you think of egusi, coconut and peanut plant milk? CLOSING (5 minutes) Thank you for your comments and your time. This has been a valuable session. Focus groups are one technique that assists market researchers in satisfying customer needs. Thank you again. 106 University of Ghana http://ugspace.ug.edu.gh APPENDIX 3 Recruitment Questionnaire for consumer tests (RPM and BIB) 1. Screener Q1. Are you willing to participate in an approximately 15-minute evaluation test on dairy and plant milk? a. Yes b. No If No, do not continue filling the questionnaire If Yes, continue to Q2 Q2. Are you allergic/intolerant to any of the following? a. Milk and dairy products b. Peanuts c. Melon seeds d. Coconuts e. Tiger nuts f. Soybeans g. Other, please start…………………………………… Terminate interview if participant has any milk allergies or is allergic to any of the ingredients. Q3. How often do you drink plant based dairy substitutes? a. Everyday b. Once a week c. Once every two weeks d. Once a month e. Less than once a month f. Never If f is selected, terminate interview. 2. Consumption Pattern Q4. How often do you consume milk and dairy products? a. Everyday b. Once a week c. Once every two weeks d. Once a month e. Less than once a month f. Never 107 University of Ghana http://ugspace.ug.edu.gh Q5. How many liters of plant milk do you consume in a week (A bottle of Vitamilk is 300ml and a pack of Soymilk or Alpro is 1000ml) …………………………… Q6. How much do you spend on plant milk in a week (in GHS) ................................. 3. Demographics Q7. Gender (NOTE GENDER, DON’T READ) c. Male d. Female Q8. What is your age range? e. 12-17 f. 18-24 g. 25-34 h. 35-44 i. 45-54 j. 55-64 k. 65+ Q9. What is your highest educational level? g. School leaver h. Junior High School i. Senior High School j. Tertiary k. Postgraduate l. Other, please specify…………………………. Q10. What is your occupation? a. Professional (Doctor, Lawyer, Engineer, Architect) b. Public servant (Accountant, Lecturer, Administrator) c. Civil servant (Nurse, teacher, uniformed works) d. Business owner e. Petty trader f. Unskilled laborers (cleaners, sweepers, watchman etc.) g. Skilled laborers (painters, mason, carpenters, seamstress/tailor etc.) h. Contractor 108 University of Ghana http://ugspace.ug.edu.gh i. In – between work j. Student k. Other, please specify…………………………………. Q11. What is your ethnicity? h. Ewe i. Akan-Forest areas (Akuapem, Akyem, Ashanti’s, Sefwi, etc.) j. Akan-Coastal areas (Fantes, Nzema, Ahanta etc.) k. Guan l. Northern m. Ga/Adangbe n. Other, please specify…………………………………… 109 University of Ghana http://ugspace.ug.edu.gh APPENDIX 4 Parental assent form for RPM Title: Optimization and Characterization of a 3-blend plant milk Beverage Principal Investigator: Dr. Maame Yaakwaah Blay Adjei 0545525974 Student Ama Frempomaa Oduro 0266394743 Address: Department of Nutrition and Food Science, School of Biological Sciences, College of Basic and Applied Science, University of Ghana, Legon, Accra. General Information about Research Your child is interested in participating in a food taste test. This will involve tasting plant and dairy milk products. During the test, your child will be given a reference product to evaluate first, after which he/she will be given 8 other products to compare with this reference product. He/she will have to simultaneously rate how much they like and how different the test product is compared to reference product. This will be done by placing the product code on a map provided to them during the test. This activity involves the use of all of their basic senses to evaluate and assess the products. Possible Risks and Discomforts In general, food taste tests are non-invasive and should not be a source of risk to your child’s health or person. Foods that are tasted are normal foods or the ingredients used to make normal foods and should not pose a risk to your child’s health unless they are allergic to milk, soybeans, melon seeds, tiger nuts, and peanuts. The beverages have been microbiologically tested and are safe for consumption. Possible Benefits By participating in this food taste test, your child will be contributing immensely to the development of other plant milk products in Ghana. This is a huge emotional benefit to your child as they will have a sense of achievement to contributing significantly to scientific research. Alternatives to Participation Should your child choose not to participate in this study or for other reasons your child is excluded from this study by the researcher, there will be other opportunities to participate in 110 University of Ghana http://ugspace.ug.edu.gh sensory evaluation research at the Department of Nutrition and Food Science, University of Ghana. If you allow us, we will contact you regarding other studies similar to this one. Confidentiality You should be assured that the data your child provides in any sensory evaluation will be kept confidential by the sensory research team. He/she will never be personally identified in any work published as a result of his/her participation in any sensory study without you and his/her prior consent. Compensation At the end of the study, your child will be given a small token to show our appreciation for his/her time spent on the project. You should understand that there is no economic benefit to you or your child for participating in a sensory study, only the emotional benefit of knowing that he/she has contributed significantly to the development and improvement in the quality of the products tasted. This benefit cannot be overlooked. Additional Cost There is no additional cost to you for your child participating in a sensory study organized by the Department of Nutrition and Food Science. Voluntary Participation and Right to Leave the Research Although we would like your child to complete any sensory project you agree for them to participate in, you should know that his/her participation is purely voluntary and he/she has the right to withdraw from a study without giving you or us any explanation and without any penalty to him/her. His/her withdrawal from a test will not negatively affect his/her personal relationship with the investigator, the department or the university as a whole. Notification of Significant New Findings To preserve the scientific quality of the data we collect in food taste tests, we are unable to disclose too much information about the products we test at the onset of the project. However, if your interest in the products is raised through your child’s participation in the project, we can provide additional information about the products to you at the end of the project. You will have to leave your details with the investigator to share such information about the products with you at the end of the study. Contacts for Additional Information For information and questions about this study in particular or any sensory evaluation test in general, please contact: 111 University of Ghana http://ugspace.ug.edu.gh Dr Maame Yaakwaah Blay Adjei, Department of Nutrition and Food Science, University of Ghana Email: myblay@ug.edu.gh Tel: 0545525974 Your Child’s Rights as a Participant This research has been reviewed and approved by the Ethics Committee for Basic and Applied Science (ECBAS). If you have any questions about your child’s rights as a research participant, you can contact the ECBAS Office through the address below Administrator, Ethics Committee for Basic and Applied Sciences College of Basic and Applied Sciences University of Ghana P. O. Box LG 68 Legon – Accra Tel: +233244692728 Email: saddo@staff.ug.edu.gh / saddo@ug.edu.gh PARENT/ GUARDIAN AGREEMENT The above document describing the benefits, risks and procedures for the sensory evaluation of foods has been explained to me. I have been given an opportunity to have any questions about the research answered to my satisfaction. I agree to let my child participate as a volunteer. _______________________ ______________________________________ Date Name and signature or mark of parent/guardian If parent/ guardian cannot read the form themselves, a witness must sign here: I was present while the benefits, risks and procedures were explained to the volunteer. All questions were answered and the volunteer has agreed to take part in the research. _________________ _______________________________________ Date Name and signature of witness I certify that the nature and purpose, the potential benefits, and possible risks associated with participating in this research have been explained to the above individual. ________________ _______________________________________ Date Name & signature of person who obtained consent 112 University of Ghana http://ugspace.ug.edu.gh APPENDIX 5 Consent form for RPM-Adult Project Title: Optimization and Characterization of a 3-blend plant milk Beverage Principal Investigator: Dr. Maame Yaakwaah Blay Adjei Student Ama Frempomaa Oduro 0266394743 Address: Sensory Evaluation Laboratory, Department of Nutrition and Food Science, School of Biological Sciences, College of Basic and Applied Science, University of Ghana, Legon, Accra. General Information about Research You have been invited to partake in a food taste test involving dairy and plant milk. During the test, you will be given a reference product to evaluate first, after which you will be given 8 other products to compare with this reference product. You have to simultaneously rate how much you like and how different the test product is compared to reference product. This will be done by placing the product code on a map provided to you during the test. This activity involves the use of all of your basic senses to evaluate and assess the products. Food taste test is an individual work activity and will involve no discussion with other participants on how you feel about the foods we present to you. In any instance, the researcher will provide you with further details on the test you are to perform and the assessment protocol you should use. Possible Risks and Discomforts In general, this consumer acceptance test, like other food taste tests, is non-invasive and should not be a source of risk to your health or person. The foods you have to taste are all normal foods or the ingredients used to make normal foods. Unless you are allergic or sensitive to dairy milk, soybeans, melon seeds, peanuts, coconut and tiger nuts, this test should not pose a risk to you. The products are handled hygienically, have passed microbial testing and are safe for consumption. If you feel uncomfortable at any point, please call the attention of the researcher who will be able to help you. Possible Benefits By participating in the food taste test, you are contributing immensely to the development of plant milk beverages from our local crops. This is a huge emotional benefit to you as you will 113 University of Ghana http://ugspace.ug.edu.gh have contributed significantly to increasing the utilization of our local crops in Ghana, leading to a boost in the local economy, improved food security and sustainability in Ghana. Confidentiality The data you provide to us will be kept confidential by the research team. You will never be personally identified in any work published as a result of your participation in any taste test without your prior consent. We will protect your personal information and not hand this to any third party. Unless you give us permission to contact you again for any sensory work we carry out at the Department of Nutrition and Food Science, we will not keep your contact information after the end of the research project. If you allow us to contact you again, we will only keep your contact details for the purpose of contacting you for sensory studies only and will not give your contact information to any third party. Your details will be kept in a secure file with the sensory research team. Compensation At the end of the study, you will be given a token to show our appreciation for your time spent on the project. You should understand that there is no economic benefit to you for participating in this sensory study, only the emotional benefit of knowing that you have contributed significantly to the development and improvement in the quality of our local foods. This benefit cannot be overlooked. Additional Cost There is no additional cost to you for participating in this sensory study organized by the Department of Nutrition and Food Science. Voluntary Participation and Right to Leave the Research Although we would like you to complete the study, you should know that your participation is purely voluntary and you have the right to withdraw from the study without giving us any explanation and without any penalty to you. Your withdrawal from the study will not negatively affect your personal relationship with the investigator, the department or the university as a whole. Termination of Participation by the Researcher It is possible that for some tests you sign up to participate in, some exclusion criteria will exempt you from participating. You will be notified of such studies at the onset. If in the middle of a test the investigator realizes that you are not capable of completing a test the investigator may ask you to discontinue the test. This does not have any negative consequence on your relationship with the investigator, the department or the university. You should understand that such decisions are made purely on the basis of preserving the scientific quality of the data we collect from our volunteering participants and have no personal bias to you. 114 University of Ghana http://ugspace.ug.edu.gh Notification of Significant New Findings To preserve the scientific quality of the data we collect in sensory testing, we are unable to disclose too much information about the products we test at the onset of the project. However, if your interest in the products is raised through your participation in the project, we can provide additional information about the products to you at the end of the project. You will have to leave your details with the investigator to share such information about the products with you at the end of the study. Contacts for Additional Information For information and questions about this study and general sensory tests and protocols at the Department of Nutrition and Food Science at the University of Ghana, please contact: Dr. Maame Yaakwaah Blay Adjei, Department of Nutrition and Food Science, University of Ghana Email: myblay@ug.edu.gh Tel: 0545525974 Your rights as a Participant This research has been reviewed and approved by the Ethics Committee for Basic and Applied Science (ECBAS). If you have any questions about your rights as a research participant, you can contact the ECBAS Office through the address below Administrator, Ethics Committee for Basic and Applied Sciences College of Basic and Applied Sciences University of Ghana P. O. Box LG 68 Legon – Accra Tel: +233244692728 Email: saddo@staff.ug.edu.gh / saddo@ug.edu.gh VOLUNTEER AGREEMENT The above document describing the benefits, risks and procedures for the sensory evaluation of foods has been read and explained to me. I have been given an opportunity to have any questions about the research answered to my satisfaction. I agree to participate as a volunteer. _______________________ _________________________________________________ Date Name and signature or mark of volunteer 115 University of Ghana http://ugspace.ug.edu.gh If volunteers cannot read the form themselves, a witness must sign here: I was present while the benefits, risks and procedures were read to the volunteer. All questions were answered and the volunteer has agreed to take part in the research. _______________________ __________________________________ Date Name and signature of witness I certify that the nature and purpose, the potential benefits, and possible risks associated with participating in this research have been explained to the above individual. _______________________ _________________________________ Date Name and Signature of Person Who Obtained Consent 116 University of Ghana http://ugspace.ug.edu.gh APPENDIX 6 Ballot Sheet for BIB CONSUMER ACCEPTANCE TEST OF PLANT MILK Panelist Code: ________________ Date: _______________ Please indicate your age range a. 18-24 b. 25-34 c. 35-44 d. 45-54 e. 55-64 f. 65+ Please select your gender (a) Male (b) Female You have been provided with coded products of plant milk. Please write the codes of the products in the spaces provided in the order in which they have been presented and evaluate them. Where tasting is involved, please cleanse your mouth with the water between products and wait for about 30s before evaluating the next product. Please indicate how much you like or dislike each of the given sensory attributes using the scale below: Scale: 1=dislike extremely 2= dislike very much 3= dislike moderately 4= dislike slightly 5= neither like nor dislike 6= like slightly 7= like moderately 8= like very much 9= like extremely Tick box with number corresponding to how much you like or dislike the product as explained above 1 2 3 4 5 6 7 8 9 117 University of Ghana http://ugspace.ug.edu.gh Product Code _________ 1. How much do you like or dislike this product OVERALL? Please tick box with number corresponding to how much you like or dislike the product as explained above 1 2 3 4 5 6 7 8 9 2. How much do you like or dislike the APPEARANCE of this product? Please tick box with number corresponding to how much you like or dislike the product as explained above 1 2 3 4 5 6 7 8 9 4. How much do you like or dislike the FLAVOUR (Aroma and Taste) of this product? Please tick box with number corresponding to how much you like or dislike the product as explained above 1 2 3 4 5 6 7 8 9 5. How much do you like or dislike the MOUTH FEEL of this product? Please tick box with number corresponding to how much you like or dislike the product as explained above 1 2 3 4 5 6 7 8 9 5. How much do you like or dislike the CONSISTENCEY (thickness/viscosity) of this product?? Please tick box with number corresponding to how much you like or dislike the product as explained above 1 2 3 4 5 6 7 8 9 6. How much do you like or dislike the AFTERTASTE of this product?? 118 University of Ghana http://ugspace.ug.edu.gh Please tick box with number corresponding to how much you like or dislike the product as explained above 1 2 3 4 5 6 7 8 9 Please comment on anything that you liked about this product ……………………………………………………………………………………………………… ……………………………………………………………………………………………………… ……………………………………………………………………………………………………… ……………………………………………………………………… Please comment on anything that you disliked about of this product ……………………………………………………………………………………………………… ……………………………………………………………………………………………………… ……………………………………………………………………………………………………… ……………………………………………………………………… 119 University of Ghana http://ugspace.ug.edu.gh APPENDIX 7 Laboratory Trial of RPM Test location The laboratory trial took place at the Sensory Evaluation Laboratory at Department of Nutrition and Food Science at University of Ghana. The laboratory is fitted with 8 individually partitioned booths, a discussion area and preparation area. It has white light and is noise and odor free. Researchers have minimal contact with assessors and samples are served through a hatch from the preparation area. The environment is temperature controlled by air conditioning and the temperature is maintained at 25oC ± 2. Participants Participants were recruited from the University of Ghana Campus, recruitment was by one-on- one interviews using a recruitment questionnaire (Recruitment questionnaire for this test is presented in appendix 2). The selection criteria were as follows; potential participants were to be consumers of plant milk and were not allergic or intolerant to milk and milk products, coconuts, cashew, almond, soybeans, rice and any of the ingredients used in the beverages. Potential participants who met the selection criteria were selected for the test. A total of 15 plant milk consumers aged between 18-34 years were selected to partake in the trial evaluation. Assessors signed a consent form and demographic data was collected from them before they started the tests. 120 University of Ghana http://ugspace.ug.edu.gh Products Table 7.1 shows the products used for the laboratory trials of RPM Table 7 1 Products used for laboratory trial Sample Product code Description 1 L1 Alpro Cashew Original 2 L2 Alpro Coconut Original 3 L3 Alpro Roasted Almond Original 4 L4 Alpro Almond touch of vanilla 5 L5 Alpro Soya Vanilla 6 L6 Alpro Rice Dolce 7 L7 VitaMilk Soy Milk UHT 8 L8 Even Full Cream milk All products used were commercial plant and dairy milk beverages bought from supermarkets in Accra. The dairy milk beverage was used as the reference. 10ml of products were served in 20ml plastic cups whilst 20ml of reference was provided. The products were pre-poured and kept in the refrigerator and assessors were served from the refrigerator in order to maintain the product temperature of 8oC-10oC. All products were served in a monadic sequential order according to a balanced design, using the William’s design in Compusense Saas (Compusense® Inc. Guelph, Ontario, Canada). Test protocol The map scale used for the RPM was a T-Map scale as shown in Figure 3.7.1. A description of how to use the t-map id found in section 3.3.3.1. The test was set up using Compusense Saas (Compusense® Inc. Guelph, Ontario, Canada), assessors did their evaluations on lab tops in individual partitioned booths. Assessors were given a verbal explanation of the test procedure 121 University of Ghana http://ugspace.ug.edu.gh after which they were given the reference product to evaluate first and then were given the plant milk products. They were instructed to simultaneously evaluate how different and how much they like or dislike the products compared to the reference. After each product, they were made to rinse their mouths with water. Figure 7.1 and 7.2 shows the liking scores and product map for the laboratory trial respectively. 40 A 35 30 B BC B 25 BC BC C 20 15 D 10 5 0 L1 L2 L3 L4 L5 L6 L7 H Samples Figure 7 1 Liking scores for laboratory trial of RPM Error bars are standard error of the means. Letters represent Fisher’s LSD post hoc analysis of means. Samples with no letters in common are significantly different at the 95% confidence level. 122 Liking compared to Reference University of Ghana http://ugspace.ug.edu.gh Biplot (axes F1 and F2: 100.00 %) L7 2 Liking 1.5 Difference 1 L5 0.5 L4 L3 0 L2 -0.5 L6 -1 L1 -1.5 L8 -2 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 F1 (62.37 %) Figure 7 2 Product map for Laboratory trial Legend: L1 ( Alpro cashew Original); L2 (Alpro coconut original): L3( Alpro Roasted Almond Original); L4 ( Alpro Almond touch of vanilla); L5 (Alpro Soya Vanilla); L6( Alpro Rice Dolce); L7 (Vitamilk Soy Milk UHT); L8 ( Even Full Cream milk Sterilized UHT) 123 F2 (37.63 %) University of Ghana http://ugspace.ug.edu.gh APPENDIX 8 Events used for RPM in social setting, pictures of RPM and ANOVA table of RPM in social and laboratory settings Table 8 1 Events used for RPM in social setting Pre-Christmas Spirit of Christmas Modality Ice Cream Fair Market Day fair 22nd and 23rd Test date 15th December, 2017 26th December, 2017 December, 2017 Truth Marketing ARK Event MBC Africa/East Organizers/Location Communication/ Management/ East Legon Alliance Francaise Legon Type of Event Pre-Christmas Sales Pre-Christmas Sales Food Fair Music Yes Yes Yes No MC, but singer There was an MC Master of Ceremony hired to sing Yes and also artistes to (MC) Christmas carols perform at the event Opportunity to sit No Yes Yes A mixture of food A mixture of food Mainly ice cream, Items sold and non-food items and non-food items pastries and drinks. Free fruit juices and Free pastries for Free Items No bissap sampling Number of vendors 8 20 15 Mainly middle to Social Class Mixed Mixed upper class People came as families (parents and People mostly young People came in social Group behavior children), others came with friends groups. No children came as groups or and socialized singles People trickled in, People came in People trickled in and Social atmosphere stayed a bit and went groups, stayed and went away away socialized 124 University of Ghana http://ugspace.ug.edu.gh Figure 8 1 Assessors using the T-map scale to evaluate plant milk in social setting RPM Figure 8 2 Assessors evaluating plant milk in social event RPM 125 University of Ghana http://ugspace.ug.edu.gh Figure 8 3 Assessors evaluating plant milk in the laboratory setting Table 8 2 ANOVA table comparing Laboratory and social setting RPM Sum of Mean Scale Type Source DF squares squares F Pr > F Model 15 115761.7542 7717.45 57.5705 < 0.0001 Error 1417 189951.9092 134.0522 Corrected Total 1432 305713.6634 Liking (Y-Axis) Samples 7 100317.2061 14331.03 1 06.9064 < 0.0001 Location 1 14554.44625 14554.45 108.573 < 0.0001 Samples*Location 7 907.1609501 129.5944 0.966746 0.453965 Table 8.2 is the ANOVA table comparing the laboratory and social setting RPM, for the dependent variable “Liking”, the p-value of the F statistic shows that the information brought by the explanatory variables is significantly better than what a basic mean would bring. The samples and Location bring significant information to explain the variability of the dependent variable “Liking”. Amongst the explanatory variables, the samples are the most influential. 126 University of Ghana http://ugspace.ug.edu.gh 127