Encontre seu próximo livro favorito

Torne'se membro hoje e leia gratuitamente por 30 dias.
Rapid Sensory Profiling Techniques: Applications in New Product Development and Consumer Research

Rapid Sensory Profiling Techniques: Applications in New Product Development and Consumer Research

Ler amostra

Rapid Sensory Profiling Techniques: Applications in New Product Development and Consumer Research

Comprimento:
1,157 página
7 horas
Lançado em:
Nov 28, 2014
ISBN:
9781782422587
Formato:
Livro

Descrição

Sensory analysis is an important tool in new product development. There has recently been significant development in the methods used to capture sensory perception of a product. Rapid Sensory Profiling Techniques provides a comprehensive review of rapid methods for sensory analysis that can be used as alternatives or complementary to conventional descriptive methods. Part one looks at the evolution of sensory perception capture methods. Part two focuses on rapid methods used to capture sensory perception, and part three covers their applications in new product development and consumer research. Finally, part four explores the applications of rapid methods in testing specific populations.
Lançado em:
Nov 28, 2014
ISBN:
9781782422587
Formato:
Livro

Relacionado a Rapid Sensory Profiling Techniques

Títulos nesta série (295)
Livros relacionados
Artigos relacionados

Amostra do Livro

Rapid Sensory Profiling Techniques - Elsevier Science

profiling

1

The use of rapid sensory methods in R&D and research: an introduction

J. Delarue    UMR1145 Ingénierie Procédés Aliments, AgroParisTech, INRA, Cnam, Massy, France

Abstract

This chapter analyses the recent evolution in the use of rapid sensory profiling techniques with respect to various situations where descriptive sensory information is needed in research, in product development, as well as in market research and consumer science. The most frequent objectives and constraints are reviewed in order to provide guidelines for users of rapid sensory methods. Implications of methodological changes in descriptive analysis are discussed. The chapter describes how the diversification of the sensory toolkit may impact on sensory activities, and why sensory scientists and their stakeholders should benefit from earlier and more integrated sensory measurements. Eventually, a SWOT analysis of rapid profiling methods is provided.

Key words

Integrated sensory measures

Sensory diversity

Profiling with consumers

Professional sensory expertise

Sensory toolkit

1.1 Introduction and context

1.1.1 Evolution in the use of descriptive analysis (DA)

Over the past decades, the use of sensory techniques has undergone drastic change. Sensory professionals in industry, as well as sensory scientists in academia, have changed the way they use these methods, and the way they use their outcomes in research and development projects. Students in sensory and food science programmes are now familiar with a series of different techniques, and the number of new methods published every year in the literature is accelerating. Among other factors, time and economic constraints that go along with industrial needs have certainly driven this evolution.

Description and quantification of human perception are difficult tasks, and sensory DA techniques are among the most sophisticated tools in the arsenal of the sensory scientist (Lawless and Heymann, 2010). Sensory profiling is often used as short-hand for sensory DA, which is in fact a name for a class of methods rather than a unique technique (Dijksterhuis and Byrne, 2005). For a review of these methods, see also Murray et al. (2001).

Sensory DA has long been a must have tool in the fast moving consumer goods (FMCG) industry, and it has proven to be extremely useful to key industrial functions such as R&D and quality management. However, probably victims of their own success, conventional profiling methods such as quantitative descriptive analysis (QDA®) (Stone et al., 1974) or the Spectrum® method (Meilgaard et al., 1999) are becoming less adaptable to growing and changing demand. In everyday industrial practice, the need for descriptive information is now very diverse, and sensory services are frequently overwhelmed with demands from stakeholders who may have multiple objectives. The cost for accuracy and reliability of conventional profiling methods is time and effort. As a result, sensory teams that rely only on conventional profiling often cannot be fast enough to meet this demand. Also, these methods are not very flexible, and certainly lack adaptability, in the face of an ever-changing market. Some companies thus question the worth of time and investment in heavily trained panels, and look for faster and cheaper methods. Varela and Ares (2012) have pointed out that the constraints in having a trained descriptive panel could also be a problem in academic research, when a short project does not justify the training of a panel from scratch, or when the lack of funding does not allow it. Besides, although conventional profiling is considered as a reference measurement in DA, a more general observation is that it is questionable whether a single method can fulfil multiple different objectives.

1.1.2 Emergence of rapid sensory profiling methods

Following this evolution, sensory practitioners have started using various methods in addition to conventional profiling to capture sensory perception for their research or in the frame of product development. Some of these methods have been known for decades, whether in sensory science or other fields, while others are new or are adaptations of older methods. They can all be seen as rapid alternatives to sensory profiling although, strictly speaking, not all of them can be called sensory profiling methods as they do not rely on description per se, i.e. when products are described in terms of sensory attributes. However, like conventional DA, they may all provide access to the relative sensory positioning of a set of products, which is most often represented in the form of a sensory map. As a result, even if it may be considered improper usage, these methods are commonly referred to as rapid sensory profiling methods. In their review of new descriptive methods, Valentin et al. (2012) have categorized these alternative methods into three classes depending on the nature of the evaluation task assigned to the panellists: methods based on verbal descriptions of products (e.g. Flash Profile (Dairou and Sieffermann, 2002), check-all-that-apply (CATA) (Adams et al., 2007)); methods based on the measurements of between-product similarity or differences (e.g. Free Sorting (Lawless, 1989; Lawless et al., 1995) and its variants, projective mapping (Risvik et al., 1994) and napping (Pagès, 2003, 2005)); and methods based on the comparison of individual products with a reference or a set of references, as in polarized sensory positioning (PSP) (Teillet et al., 2010). Dynamic methods, such as temporal dominance of sensations (TDS) (Pineau et al., 2009), could form a fourth category, although TDS is also based on the use of attributes.

Interestingly, the increasing use of these new methods has been conducive to enlarging the scope of traditional descriptive sensory analysis, as is discussed later in this chapter and amply illustrated throughout this book. Thus, it is worth underscoring that these rapid sensory profiling techniques are not only rapid alternatives to conventional sensory profiling ones, but that they may also be used in situations where sensory profiling had not been previously possible.

1.1.3 Aims and needs

Today, the circumstances under which DA is needed are very diverse, and range from daily industrial practice to academic research, and from quality control to marketing studies. They differ in terms of both objectives and constraints. As a result, it would be inappropriate to rely on just one technique to fit all situations. Therefore, before going through the various sensory techniques presented in this book, it will be useful to review the different uses of DA with respect to these situations and the related objectives. Accordingly, it is important to analyse the major consequences in terms of requirements (what is needed) and constraints (applicability), as these considerations will generally guide the choice of a DA method. What is more, as noted by Dijksterhuis and Byrne (2005), the objectives (the research question) may also influence the methodological choices at different steps of a given sensory profiling study.

It should first be noted that quality control is outside the scope of this book, although it is an important field of application for sensory analysis. Rapid profiling methods are indeed not developed for the purpose of monitoring product changes over time, variations in raw materials or production, which require measurement tools that may take time to implement but which can be used with confidence in the long run. That is quite opposite to rapid profiling methods. Of course, one could imagine adapting some of these methods so that they could be used in a quality control context, but conventional sensory profiling or other dedicated methods would seem to be much more adapted. The reader interested in such use of sensory measurements is thus invited to refer, for instance, to Muñoz (2002) and Costell (2002).

1.1.3.1 Use of DA in R&D projects

Most R&D projects in the FMCG industry aim at developing new products, optimizing existing products, or maintaining product properties with respect to other constraints (e.g. supply issues, regulation changes, cost reduction, packaging, sustainability, etc.). In such projects, developers need to take into account the sensory properties of the products they are working on, even when they are not the primary focus of the project. The way the sensory properties are taken into account may, however, take different forms that in fact reflect different objectives. The most frequent objectives of DA are:

• To compare products with existing products on the market, to compare a new or an improved product to previous versions, or to compare several prototypes;

• To improve or to optimize. Most frequently, optimization is understood in terms of liking (but not limited to it). Therefore, it implies that one has the ability to relate the sensory properties of the product to liking, in order to identify drivers of preference and, ultimately, to determine optimal sensory properties;

• To understand how product formulae or process variables affect its sensory properties. An underlying objective is usually to acquire knowledge and develop technical skills and expertise;

• To communicate with others: other team members, other teams (marketing, suppliers and customers, subsidiaries, local teams, etc.).

One could easily imagine that, depending on these objectives, the needs for sensory information will differ. Sometimes, information on how products can be sensorily grouped or positioned relative to each other would be sufficient, while in other cases an accurate quantification with finely tuned attributes would be needed.

1.1.3.2 Use of DA in marketing and consumer research

Obviously, sensory DA aims at measuring product characteristics. However, it is now widely assumed that sensory measurements are subjective, and essentially capture a product × subject interaction (i.e. the characteristics being perceived by the subjects) rather than absolute properties of the products. Naturally, it does not mean that sensory measurements are not sound and reliable, but a direct consequence of this subjectivity is the acknowledgement of inter-individual differences as the foundation of sensory measurements. Accordingly, many of the descriptive techniques that have been developed over the past 10–15 years allow for taking into account interindividual differences in perception. This is naturally conducive to seeing the sensory panel not only as a sensory measure instrument, but also as a sample of subjects that may be representative of consumers’ perceptions. That approach to sensory measurement has been coined Sensory Evaluation II by O’Mahony (1995).

This point becomes crucial in a competitive context where FMCG companies pay more and more attention to the way consumers perceive their products. In this respect, R&D and market research share common objectives. Consequently, there is a very strong trend toward more consumer-orientated approaches to product innovation and development (MacFie, 2007; Jaeger and MacFie, 2010), which implies consumer-orientated measurements, including for descriptive purposes (van Kleef et al., 2005; Tuorila and Monteleone, 2009). Many sensory professionals have thus accepted to somehow compromise on the accuracy of DA in favour of consumer-orientated results.

Dijksterhuis and Byrne (2005), however, have called into question the use of conventional sensory profiling studies in the perspective of studying consumer behaviour. They argue that the cognitive processes involved in (attribute-based) profiling require conscious action by the panellists, and hence they are required to enter into an analytical mode of thinking that may lead to results that are somewhat different from the global, and sometimes unconscious, sensory image that forms in our minds when normally consuming or using a product. An implicit assumption in sensory profiling is, indeed, that perception can be split up into separate attributes, which may sometimes be challenging (Lawless, 1999; Murray et al., 2001).

Another major limitation to performing DA with consumers is the use of a common language and the associated training required to align concepts (O’Mahony, 1991). Seeking a common language may lead to overlooking the inter-individual diversity of consumers in terms of their descriptive language and in their perception. Removing the semantic constraints through free choice of attributes, descriptive words or semantic-free tasks has allowed sensory scientists to conduct DA more easily with consumers (Thomson and McEwan, 1988; Gains and Thomson, 1990; Jack and Piggott, 1992; Veinand et al., 2011). Beside methods that do not rely on a common vocabulary, the emergence of non-verbal and holistic methods has expanded the possibilities of capturing consumers’ sensory perceptions (Risvik et al., 1994; Faye et al., 2004; Moussaoui and Varela, 2010).

Overall, more attention is now paid to the variety of consumers’ perceptions and judgements. The main challenge for sensory scientists is then to deal with this variety and to provide efficient solutions to stakeholders. In some cases, this implies investigating a possible typology of the consumers in terms of perception, not just in terms of preferences (see Chapter 6 and Chapter 20 for examples of such approaches).

In addition to this, a common feature of rapid sensory profiling methods is the absence of (or very limited) training. This has further heightened interest in the use of these methods in consumer studies. As a result, as Varela and Ares (2012) have underlined, the line is now blurred between sensory and consumer science, and rapid sensory profiling methods are now increasingly used to capture consumers’ perceptions.

In the same line of thinking, market researchers have attempted to use rapid sensory methods to evaluate how consumers perceive products, not only in terms of sensory attributes, but also in terms of expectations, emotions, evocations, lifestyles, etc. Ballay et al. present a very nice example of how Flash Profile can be adapted to investigate these aspects in Chapter 19 of this book. This evolution has also led researchers to adapt sensory techniques to investigate concept fit (Carr et al., 2001; Lee and O’Mahony, 2005) and conceptual associations (see Chapter 5 of this book).

The expanding use of rapid sensory methods in this direction may also be seen as an aid in better communicating sensory properties to consumers, either in a direct (verbally using claims, cobweb plots, etc.) or indirect manner (using sensory marketing through colours, shapes, etc.).

1.1.3.3 Use of DA in research

Naturally, DA is also used in research for various purposes, ranging from the most applied product-related issues (especially in food science) to the more fundamental psychophysics. Below are the main categories of research objectives for which DA may be involved:

• To understand how product properties affect perception (e.g. study of food-flavour interactions);

• To link sensory properties with instrumental measurements;

• To understand human perception of sensory stimuli and its link to the determinants of consumption behaviour;

• To understand inter-individual differences in perception (genetic, cross-cultural differences, etc.).

From a practical point of view it must be noted that, for all these categories of research objectives, rapid methods may be useful, especially when experimental constraints predominate (e.g. use of custom-made product tailored for research purposes, use of short shelf life samples, short time availability with the subjects, or the need to interview large samples of subjects in a given population). In addition to this, it is worth noting that rapid sensory methods can also be used with subjects with limited cognitive abilities, because these methods usually involve simpler tasks. Examples of such uses are discussed in Chapters 22 and 23.

From a conceptual point of view, it must be noted that thanks to the generalized use of multivariate analysis techniques, researchers in sensory science have started to pay more attention to the relative positioning of the objects rather than to product scores on separate attributes. As a result, rather than measuring the stimulus by conventional physical means as a psychophysicist might do, one might create a multidimensional space, and then use the coordinates of that space as a surrogate set of physical measures made on the same stimuli in the test set (Moskowitz, 2003). The increasing use of rapid sensory profiling methods in research is an integral part of how that evolution has taken place (Delarue et al., 2004). This has indeed been conducive to creating a more holistic picture of response to stimuli. As described by Moskowitz (2003), this type of thinking represents a psychophysical mindset (functional relations between variables) applied to new types of data (locations of products using a multidimensional coordinate space).

Last but not least is the research in sensory methodology and in sensometrics that is actively contributing to the evolution of sensory science. Perhaps the main issue for this methodological research is to keep contributing to the development of new methods and adaptation of existing methods with due regard to the objectives and constraints for potential users of these methods.

1.2 Methodological evolution

Along with the evolution of sensory techniques, the greater availability of data analysis techniques and statistical software has greatly contributed to the increasingly widespread use of rapid sensory methods. Sophisticated multivariate techniques such as generalized procrustes analysis (GPA), multiple factorial analysis (MFA) and multidimensional scaling (MDS), which allow analysis of more complex datasets provided by rapid techniques, are now available to anyone. As a result, it is possible for any well-trained sensory practitioner to analyse data with appropriate techniques and hence to use rapid profiling methods, without it being a hurdle. This has obviously resulted in the diversification of the toolkit for the sensory practitioner, who may now choose which method to apply depending on the type of information that is needed and on the metrological properties of each method, but also according to criteria of rapidity and flexibility. In addition to this, panel training being a lesser requirement with rapid sensory profiling methods, new types of subjects can be used for conducting DA, which opens additional perspectives.

1.2.1 Rapidity

It must be stressed that not all the methods presented in the book are rapid per se. Rapidity here is indeed a relative notion, and is in fact multifaceted. Many of the so-called rapid profiling methods are seen as rapid because they do not require an extensive training phase. As a matter of fact, most of them do not require training at all. As a result, the time necessary to acquire sensory data starting from scratch is indeed much shorter than with any conventional profiling method. However, this may only be true in the absence of prior information or of a trained panel or, as Murray et al. (2001) observed, that while, on the one hand, time constraints may be limiting in their profiling methods based on extensive training, on the other hand they may be considered to be timesaving in the long run.

Actually, when considering the time needed for subjects to complete the evaluation task, the cards are being reshuffled. Although the simplest comparative tasks, as in Free Sorting, may still be faster, monadic sequential evaluation of the samples in conventional profiling is not much longer than other evaluation tasks. In some cases, this is in fact quite the opposite; this would be the case for instance with flash profiling if we consider the length of the evaluation session.

It may also be worth considering other parameters, such as the time to recruit subjects and to set up a panel, the time spent by the experimenter to collect and organize data, and the time needed for data analysis and interpretation.

Eventually, it should be stressed that rapid does not necessarily mean instantaneous. Rather, one should consider rapid profiling methods as ways to acquire data more rapidly than with their conventional equivalents. In Chapter 20, Blumenthal and Herbeth report that in order to evaluate sensations while driving, each assessor in their study drove a total of more than 25 h and 600 km, even if they used a quicker approach than conventional profiling. The same goes with other methods: TDS is not as rapid, but it is much faster than traditional time–intensity measurements; the Ideal Profile Method may also take some time for the panellists, but it is clearly more rapid than a full external preference mapping study.

1.2.2 Flexibility

A key point in the application of rapid sensory profiling methods is that they do not require an important initial investment, since there is no need to train and to maintain a panel. There is thus no need to mobilize substantial resources to run a sensory profile (except when a large consumer sample is to be recruited). In a way, those methods thus democratize access to sensory information.

Accordingly, practitioners have started to realize that rapidity comes along with flexibility, which is perhaps an even more interesting feature. This has opened ways for the use of DA in situations to which it was previously not adapted, or even not possible to apply. These situations would notably include: sensory profiling of products with supply or sampling constraints (short shelf life, prototypes, etc.); earlier use of sensory profiling in product development or in research projects; sensory profiling in small facilities; sensory profiling with subjects who are usually not available for repeated training sessions. Flexibility would also allow using sensory information in a more interactive way, as for example for the selection of sensory-relevant variables or sensory-relevant product sets in the frame of a project. In general, having such flexible tools at hand may facilitate decision making and project management without spending too much time.

1.2.3 Diversification of the sensory toolkit

The evolution of sensory methods, and more specifically the development of rapid sensory profiling techniques, has resulted in a broad variety of available methods. The sensory toolkit is now much more complete than even a few years ago and offers several alternative solutions to the sensory practitioner to address many different objectives. When taking one step back, one become aware of the impressive number of options that underlie DA methods (Fig. 1.1). These methodological options range from the type of subjects that are included in the panel to data handling and data analysis issues, and include the different cognitive tasks, the type of measure and the nature of the attributes, the assessment mode, the type of vocabulary, the means of quantification, the context of evaluation and the design of evaluation sessions. Some of these topics have given rise to literature reviews and to further methodological development, but there are still many unexplored options and combinations that may allow addressing new goals in the future.

Figure 1.1 Analysis of the spectrum of methodological options for sensory descriptive analysis.

1.2.4 Use of DA with different types of subjects

A major consequence of the developments in DA methods is linked to the typology of subjects that are used to perform the descriptive task. In effect, descriptive sensory tasks are traditionally performed with panels of subjects with varying degrees of training or expertise, depending on the method. Although a minimal training is considered to be necessary for a series of reasons (need for a semantic consensus, attribute understanding and quantification, use of scales, repeatability, etc.), sensory scientists have found ways to get round some of these requirements in order to reduce training. The reduction or even the absence of training may imply compromising on the quality of the measurement (in terms of accuracy, precision or repeatability) or may induce more complex data interpretation. Nevertheless, many methodological developments have paved the way for running DA with subjects other than trained panellists. To date, most emphasis has been placed on the consumers versus experts paradigm (see for example Scriven (2005) for a discussion of these two types of panels), but there are certainly more options as will be presented in the following sections.

1.2.4.1 DA with consumers

As already discussed, getting descriptive information directly from consumers has long been very tempting to sensory scientists (Guy et al., 1989; Jack and Piggott, 1992). However, the custom in sensory science is to interview consumers for hedonic testing only. Accordingly, it is generally recommended to avoid descriptive or analytical questions when running a hedonic test, in order to limit potential biases (Lawless and Heymann, 2010; Prescott et al., 2011). All the more reason for not training consumers and hence risking biasing their perceptions and responses, not to mention the fact that training consumers would certainly be too time-consuming within the course of a consumer survey. The ability of untrained consumers to give reliable DA results could thus be questioned, although this ability seems to be largely underestimated (Husson et al., 2001; Worch et al., 2010). Interestingly, QDA was initially designed as a way to capture consumers’ perceptions by eliciting consumer-orientated vocabulary from the panellists under supervision of a panel administrator (Stone et al., 1974). However, recruiting consumers and training them into no-longer-naïve subjects would indeed lose much of its interest. In common practice, QDA subjects are therefore most frequently referred to as trained panellists, and consumer panels are dedicated to hedonic testing.

Yet conducting DA with consumers is extremely appealing, as it is a potential way to assess consumers’ diversity in their perception of products, while this cannot be achieved using trained panellists (because of the training and of the limited panel size). Alternately, qualitative techniques that are traditionally used in market research would only partially render this diversity, and would fail to provide actionable or quantitative data.

Since then, the number of DA studies with consumers has increased impressively. Recently, several comparison studies have shown that running such evaluations is perfectly feasible, and that the results compare very well with DA conducted with experts (Moussaoui and Varela, 2010; Worch et al., 2010). This tends to indicate that trained panellists yield sensory information that is not distorted compared to that obtained with consumers, hence supporting the use of DA with trained panels with respect to external validity. It should be noted, however, that a meta-analysis of this issue is difficult because usually there is relatively little information regarding the recruitment of participants in most published studies. For instance, the term consumers is frequently used as a synonym for untrained subjects, which may result in biased conclusions (Schutz, 1999; Scriven, 2005). The same applies to trained panellists and experts: the degree of training and/or level expertise is often not specified. Sadly, a very good illustration of this is the use of students as panellists. Using students is not intrinsically a bad thing (the present author has conducted many studies with students), but it is striking to see that they may alternatively be considered either as experts or as consumers, with apparently not much difference in their training or prior sensory knowledge. Nevertheless, conducting DA with consumers draws considerable interest. This way to use DA indeed opens perspectives to sensory science for understanding consumer perceptions, and relates more closely to consumer insights. In particular, the use of this type of analysis for investigating the diversity of consumers’ perceptions is very promising.

How many consumers should be recruited for conducting a DA is a frequent question. Although not well defined, the most frequent panel size in published studies is of about 40–50 participants, but this may vary from one study to another, from 20 to several hundred. The appropriate sample size will probably depend on the objectives of the study. Obviously, the more consumers that are interviewed, the better the products will be statistically discriminated. The logic behind hedonic testing may apply. Greater accuracy would be obtained with larger consumer samples. According to most standards, such reasoning would imply running these tests with at least 100 consumers. Besides, increasing the number of consumers would automatically increase the amount of sensory information collected. In this respect, one may refer to qualitative studies, where researchers generally use saturation as a guiding principle during their data collection. Saturation occurs when the collection of new data does not shed any further light on the issue under investigation (Glaser and Strauss, 1967). To our knowledge, no study has addressed this point in the perspective of consumer-based DA.

Another consideration is the possibility to explore consumer’ diversity, which may only be possible when a large number of consumers participate in the test or when careful segmented sampling is applied. Eventually, we should also keep in mind that the number of consumers recruited to participate in such tests greatly impacts the time and the cost of the test. This may be a major limitation to the development of such studies.

1.2.4.2 Professional sensory experts

Professionals who work with the product every day develop a sensory expertise that could be the core of their activity (e.g. perfumers), or that may be less conscious and yet extremely useful in their work (e.g. plant operators, hairdressers). In many cases, it would potentially be very interesting to include the input from those professional experts in sensory studies. Recently, researchers have started to apply rapid sensory profiling methods as a way to get sensory input from professionals. Such attempts are as yet very infrequent, but this trend is very promising, with many options remaining to be

Você chegou ao final desta amostra. Inscreva-se para ler mais!
Página 1 de 1

Análises

O que as pessoas pensam sobre Rapid Sensory Profiling Techniques

0
0 avaliações / 0 Análises
O que você acha?
Classificação: 0 de 5 estrelas

Avaliações de leitores