Application of sensory evaluation in food research

Abstract

Eating and drinking should be pleasurable. The sensory experiences evoked by foods and beverages are key to the delivery of pleasure and crucial to commercial success. Advertising and branding motivate consumers to try products, but if they don’t like a product’s appearance, aroma, flavour or texture, they will not buy it again. Specialised research techniques are available to measure, understand and optimise consumers’ sensory experiences, so that products can be designed and marketed to meet consumers’ sensory needs, thus reducing the risk of product failure. This scientific field is known as sensory evaluation. At a time when the application of sensory evaluation is growing, it is appropriate that the IJFST has dedicated an issue to papers in this field. The Professional Food Sensory Group (PFSG) of The Institute of Food Science and Technology (IFST) are pleased to be invited to provide this editorial (with no formal involvement in selecting and reviewing papers) and, thereby, contribute to greater awareness and understanding of the field. The PFSG is the body through which the IFST promotes sensory evaluation. It aims to: Establish, maintain and enhance the professional status of food sensory scientists. Facilitate better communication of food sensory matters. Provide objective scientific comment on food sensory issues. Provide guidance on food sensory practices, such as ethics, laboratory practices and relevant legislation. Assess, monitor and review courses designed to enable professional recognition of food sensory scientists. Provide career advice for food sensory scientists. Encourage young food sensory scientists into the profession and assist the professional development of food sensory scientists through membership of IFST and PFSG. The PFSG is one of the first organisations in the world to develop an accreditation scheme for sensory evaluation training to introduce consistent standards and proper guidance on the quality and content of basic sensory training courses. It is also one of the first organisations to set ethical and professional practices for the sensory analysis of foods. It regularly hosts conferences, symposia and workshops on sensory evaluation and sits on the British Standards Institute sensory committee. (Further information on the PFSG and how to join, can be found at http://www.IFST.org.uk.) Sensory evaluation is traditionally defined as a scientific method used to evoke, measure, analyse and interpret those responses to products as perceived through the senses of sight, smell, touch, taste and hearing (Stone and Sidel, 1993). It can be divided into two areas: objective (analytic) and subjective (hedonic). In objective testing, the sensory attributes of a product are evaluated by a selected and trained panel. In subjective testing, the reactions of consumers to the sensory properties of products are measured. The power of sensory evaluation is realised when these two elements are combined to reveal insights into the way in which sensory properties drive consumer acceptance. Linking sensory properties to physical, chemical, formulation and/or process variables then enables the product to be designed to deliver optimum or appropriate consumer quality, functional and emotional benefits. The science of sensory evaluation emerged in the 1940’s from its origins in quality assessments made by expert tasters employed in industries such as tea, coffee and cheese. It is now integrated into the entire product life cycle with applications in: Product development, including determining preferences, identifying sensory drivers of liking, targeting sensory-based consumer segments, competitor analysis, new concept development, product design and optimisation, scale-up and cost reduction. QA/QC, including raw material quality, sensory specifications to ensure acceptability, taint testing, shelf-life and quality auditing through supply chain. Marketing, including claim substantiation, delivering emotional benefits, and synergy with brand communication and advertising. Research to improve fundamental understanding of consumer behaviour and perception. The papers in this special edition of IJFST illustrate some of the broad range of areas in which sensory evaluation can be applied in research in the food industry, including ingredients and formulation, quality, storage and shelf-life, and health and nutrition. In the area of ingredients and formulation: An investigation of the effect of pectin concentration on processed cheese showed that pectin can be increased without damaging appearance or flavour (Macku et al.). Gamlath investigated a novel use for over-ripe bananas in flour to produce acceptable extruded snack products, measured using on a 1–9 Liking Scale with thirty consumers, that showed potential for further investigation. Wong et al. investigated sensory aroma, using descriptive analysis with eleven trained panellists, from Maillard reaction of amino acids with glucose in acidic conditions to predict flavour characteristics and found that the Maillard reaction can be used as a basis for production of specific flavours. Galmarini et al. found that matrix composition determines the aromatic profile, measured using descriptive analysis with twelve trained panellists, of spray-dried encapsulated orange flavours. In the area of quality: Treatment with hydrogen peroxide was found to increase whitening (measured using instrumental methods) of common carp fillets to be used in surimi production (Jafarpour et al.). The role of volatile compounds on aroma and flavour of coloured raw carrot genotypes was investigated using descriptive analysis with ten screened, trained panellists and instrumental techniques. A sensory map was developed for coloured carrots related to genotype (Kreutzmann et al.). Consumer attitude and behaviour towards tomatoes after 10 years of using the Belgian Flandria quality labelling system on fruit and vegetables was investigated using a survey with 373 consumers and a change in consumer attitudes was found (Verbeke et al.). In the areas of shelf-life and storage: An investigation of the effect of packaging films and storage temperature on minimally processed cauliflower showed that cauliflower maintained acceptable appearance in all the films studied (Simon et al.). Sensory assessment of shelf-life of butterhead lettuce leaves in active- and passive-modified atmosphere packages found that low storage temperatures and active atmosphere increased acceptable shelf-life, as determined by consumer purchase intent (Ares et al.). Soluble CO2 stabilisation prior to vacuum packaging of sea bream and sea bass was found to give longer shelf-life (Mendes & Goncalves). A quality and shelf-life assessment of whole and gutted sardines stored in ice was carried out (Erkan & Ozden). Exposure to ozone reduced microbes on flaked red pepper, with only a small change in acceptability (Akbas & Ozdemir). Freeze-thaw cycles were found to have a detrimental effect on ready-to-bake frozen chapaties, which could be improved by adding glycerol (Yadav et al.). In the area of nutrition and health: The addition of gums to noodles to make them gluten free, and thereby suitable for consumers with celiac disease, showed promise for further investigation (Yalcin & Basman). Fortification of wheat bread with defatted soy flour was found to give as good a loaf as a 100% wheat bread with higher nutritional quality and good consumer acceptability, measured on a 1–9 Acceptabilty Scale with 145 consumers (Mashayekh et al.). A characterisation of higher-fibre flour, dough and pasta produced using toasted durum whole meal to partially replace semolina was carried out (Baiano et al.). Flours with a higher fibre content from inclusion of king-palm residue produced acceptable cookies, with consumer response measured using a 1–9 Liking Scale and 1–5 Purchase Intent Scale with 100 consumers (Vieira et al.). Acceptable snack products were produced from highly nutritious novel raw materials amaranth and chickpea mixed with bovine lung, using added monosodium glutamate and disodium inosinate to improve texture (Cassar et al.). Sensory and physico-chemical optimisation of ice cream formulated using new sweeteners to produce a low GI Index product resulted in an acceptable candidate formulation (Whelan et al.). An physical and sensory investigation of the addition of pectin to improve texture of low fat cheese yielded promising results for further investigation (Liu et al.). The addition of probiotics to coconut flan, a traditional Brazilian dessert, showed no significant difference in acceptability over 21 days storage (Correa et al.). Consumer perceptions and attitudes to probiotic foods in Rio de Janeiro, Brazil, measured using a survey with 420 consumers, revealed that the majority of consumers could not correctly define probiotic, indicating the need for education to embed probiotic concepts (Viana et al.). It is not possible to review all the papers in detail here, but there are a few that warrant special attention for highlighting noteworthy points in sensory evaluation. Two papers demonstrate that, despite advances in instrumental analysis, sensory evaluation remains the most sensitive and reliable measurement technique in some situations. Galmarini et al. found that sensory analysis was more sensitive than the e-nose for analysing orange flavours and provided more information, in that it can be used to assess pungency and provides names for the sensations measured. Mendes & Goncalves found that sensory evaluation was the most reliable indicator of quality in sea bream and sea bass, whereas chemical indicators, except for pH, were found to be poor indicators of change in quality. Consumer attitudes and behaviours change over time because of many factors, such as availability of new information, new product introductions, changes in life-style, economic trends, etc. Verbeke et al. provide an example of such a change. They assessed consumer attitudes and behaviours to tomatoes after 10 years of using the Belgian Flandria quality labelling system. Over time, the Flandria label has become the norm, appearing on many fruits and vegetables, and so is now associated with a normal, standard tomato, giving little perceived differentiation in quality, apart from origin and production. Results from consumer research, such as liking, preference, purchase intent, attitudes, beliefs, habits, etc., need to be reassessed regularly at appropriate time intervals or when a change occurs that impacts on the commercial environment. It also means that systems based on consumer data, such as quality and grading schemes, must be checked regularly to ensure they remain consumer-relevant. The full potential of sensory evaluation is realised when sensory, consumer and and/or instrumental analyses are combined. Many of the papers in this edition use instrumental, sensory and/or consumer data as independent pieces of evidence. A few papers stand out as examples of statistically combining data from different sources to give more information. Ares et al. investigated the shelf-life of butterhead lettuce leaves in active- and passive-modified atmosphere packages. As part of the study, a sensory panel identified the defects in lettuce leaf appearance that best differentiated fresh and stored samples. They were trained on these defects and then carried out an objective quantitative descriptive assessment on lettuce leaves stored in different packaging conditions at each storage time. A separate group of forty consumers of lettuce were presented with the lettuce leave samples as above and asked a purchase intent question: ‘Imagine you are in a supermarket, you want to buy a package of minimally-processed lettuce, and you find a package of lettuce with leaves like this, would you normally buy it?’ to which they must answer yes or no. The consumer purchase intent data were linked to the objective sensory panel attributes (defects) using regression analysis. This was used to estimate the level of sensory defects with the most rapid onset, at which 25% consumer rejection would occur. The estimated levels were considerably more conservative than the arbitrary figure typically used in shelf-life studies on fruit and vegetables. In the words of the authors: ‘Results showed the importance of performing consumer studies in order to establish proper criteria to estimate the shelf life of fresh fruit and vegetables’. Kreutzmann et al. combined sensory and chemical data using principal component analysis (PCA) to understand how volatile compounds relate to aroma and flavour of coloured raw carrot genotypes. The PCA was represented as a bi-plot, which served as a visual aid to help interpretation. Correlations enabled compounds and sensory attributes to be linked. The presence of terpenes was found to be related to harsh flavour attributes. Interactions between sugars and volatiles affected sensory perception, so that the increasing sugar content resulted in masking of harsh flavour qualities. Reliable and meaningful sensory data can only be produced if the appropriate procedures are used. The paper by Galmarini et al. illustrates many of these procedures for objective sensory quantitative description, such as using an appropriate number of panellists, training the panel on attributes and scale use, using physical references to illustrate attributes, presenting samples in an unbiased fashion, assessing samples in at least duplicate (Kreutzmann et al. assessed in triplicate) and checking sensory data to ensure results are due to product differences rather than unreliable panel performance. In addition, it is good practice to screen for and select panellists with appropriate sensory abilities prior to using them in a study, as applied by Kreutzmann et al. Sensory evaluation has evolved into a complex, multi-disciplinary field that requires a high degree of scientific knowledge and skill to carry out successfully. A key challenge is to use a highly variable human ‘measuring instrument’ to gather robust, unbiased data. Even a seemingly ‘simple’ triangle test is fraught with potential pitfalls. Sensory analysis is rarely routine and often needs to be tailor made to match the product, situation or assessor. There remains a tendency for companies to hire untrained staff to run sensory programmes, sometimes driven by a shortage of experienced sensory scientists, and for scientists from other fields to attempt to carry out sensory testing without consulting a skilled sensory professional and/or statistician trained in sensory techniques (sensometrician). IJFST often finds it necessary to reject papers with a sensory element because fundamental errors have been made. The following are some of the basic elements necessary to carry out sound sensory research: The study must be conducted safely and ethically. For guidance see ‘Ethical and Professional Practices for the Sensory Analysis of Foods’ published by the IFST PFSG. Considerations include: Safe and ethical treatment of assessors (panellists, consumers and subjects): Short- and long-term health effects on assessor groups (e.g. existing medical conditions, allergies, effects of long-term product usage, etc.) Ethical treatment (e.g. informed consent, special considerations for children, etc.) Safe test materials: Safety of ingredients, (e.g. approved for use in country of test, origin, recommended intakes, allergenic effects, identification and safety assessment on taint compounds prior to tasting, etc.) Effect of processing (e.g. creation of side-products, change in nature, etc.) Safety of finished product (e.g. microbiologically safe, no leeching of harmful packaging material into product, etc.) Safe experimental design and techniques: Safety of test procedure (e.g. differences from normal life conditions, safe and hygienic test sample preparation, etc.) Test objectives must be pre-defined and an appropriate experimental design should be used to meet these objectives. A typical error is not using a formal experimental design to systematically assess effects of treatments. The correct technique(s) to meet objectives should be used, including: Appropriate test type. Typical errors include applying one technique, often descriptive analysis, in all situations, and using sample labelling that will bias response such as ‘1, 2, 3’ or ‘A, B, C’. Appropriate experimental design for serving order. A typical error is presenting samples in the same order for all assessors. Sufficient numbers of assessors must be used to meet the statistical power requirements of the sensory test. The correct type of assessors should be used. A classic error is using a trained sensory panel or quality panel to give consumer liking, pleasantness, preference or acceptability ratings. (Prediction of consumer response using a trained panel is possible, but only if a relationship has been statistically established previously.) Another error is failing to test and/or target the appropriate consumer group. Appropriate instruction on how to perform the assessment should be given to assessors. Typical errors include presenting scales with no information on what the scale is measuring or how the scale should be used, and assuming that allowing familiarisation with a product or test procedure is the same as providing training. Replication should be used where appropriate, such as for descriptive analysis. As in other scientific disciplines, replication reduces the chance of a random or indiosyncratic result, increases reliability and increases the statistical power of a test. It enables additional statistical analysis to be carried out, such as checking panel performance. Appropriate statistical analysis techniques must be applied to the data. Typical errors include using means when modes and medians are appropriate for the data, such as data generated from category scales; using multiple comparison tests without first carrying out an analysis of variance (anova) or when the anova did not show a significant difference; inappropriate application of Fishers LSD test; carrying out multidimensional scaling with an insufficient number of samples. The correct conclusions must be drawn from the data. Typical errors are to conclude that products are/are not significantly different from the results of a liking or preference test (rather than a sensory rating or difference test), and assuming that a sensory difference has been caused by a particular factor when no evidence is present. Sufficient information on how the sensory test was carried out must be reported. As in other scientific fields, this must be enough information to enable the study to be replicated. For example, how samples were prepared (cooking, serving size, etc.) and presented (serving vessel, labelling, temperature, lighting, etc.), experimental design (serving order, blocking, etc.), type of assessors (gender, age, level of training, etc.), etc. Sensory evaluation is a growing, dynamic field. It continues to broaden its applications from its roots in food and beverages to include categories as diverse as personal care products, household products, cars, mobile phones and environments, to name but a few. The role that sensory evaluation plays in organisations continues to grow. Sensory departments were typically a part of the QA function supplying data as a service. They are increasingly becoming fully integrated in all stages of a product’s life cycle from product conception to post-launch monitoring and work as partners with R&D and marketing to provide insights to help guide development and commercial strategy. As many brands and products are now global, sensory projects are increasingly required to have a global perspective. Indeed, the structure of the professional bodies in the sensory evaluation field is changing to become more global. Sensory research is delivering increasingly sophisticated techniques and better understanding of consumer perception and behaviour using a multi-disciplinary approach by linking with fields such as physico-chemistry, psychophysics, psychology, physiology, neuroscience and genomics. Some challenges remain. Education continues to be a need as a result of skills shortage in the field. Some traditional techniques are perceived as slow compared with the increasing speed of product launches, and quicker methods are under development. Testing with human subjects requires increasing attention to health, safety and ethical considerations, coupled with an increased risk of litigation. Changes in employment law in the EU are making it increasingly complex to employ sensory assessors. The future for sensory evaluation is bright, however, as it transitions from an option to a necessity to meet consumers’ sensory needs and deliver competitive advantage in the quest for successful products.