Consumer perceptions of satiating and meal replacement bars, built up from cues in packaging information, health claims and nutritional claims

Food Research International 64 (2014) 456–464

Contents lists available at ScienceDirect

Food Research International journal homepage: www.elsevier.com/locate/foodres

Consumer perceptions of satiating and meal replacement bars, built up from cues in packaging information, health claims and nutritional claims Marcelo Miraballes a, Susana Fiszman b, Adriana Gámbaro a, Paula Varela b,c,⁎ a b c

Sensory Evaluation Section, Department of Food Science and Technology, Faculty of Chemistry, University of the Republic, General Flores 2124, CP 11800 Montevideo, Uruguay Institute of Agrochemistry and Food Technology (CSIC), Agustín Escardino 7, 46980 Paterna, Valencia, Spain Nofima AS, P.O. Box 210, 1431 Ås, Norway

a r t i c l e

i n f o

Article history: Received 28 April 2014 Accepted 20 July 2014 Available online 28 July 2014 Keywords: Consumers Satiety Meal replacement Projective mapping Health claims Nutritional claims Sensory

a b s t r a c t Previous research has shown that consumers' sensory and hedonic perceptions could be greatly influenced by the messages highlighted on the front of the packaging, particularly nutrition and health claims for reduced-calorie or functional foods. In food products designed for hunger control, mentioning satiating effects or meal replacement could potentially influence the expected satiety or other perceptions, including those associated with a filling effect. This study investigated the effect of packaging information on consumer perceptions of eleven commercial chocolate-flavoured satiating and meal replacement bars. Projective mapping (Napping®) with a descriptive step was used to profile the expectations elicited by the eleven bars, based on their extrinsic characteristics (without tasting them), with two different groups of consumers in two different scenarios: mapping the samples' packaging and mapping cards with an identical design showing the most important information displayed on the packaging (product name, nutrition information panel and nutrition and health messages or claims). The terms and comments obtained in the two scenarios were analysed and perceptual spaces were generated from the sample location and attributes obtained in the Napping® exercises, using multi factor analysis (MFA). The results showed that when evaluating the real packaging, the consumers mainly focused on the meal replacement- or satiation-related messages and on the images to build up their perception, attaching less or no importance to the nutritional information that was also displayed on the box. When examining the packaging, sensory expectations – which probably emerged from the images of the bars – were also a very important factor in placing the samples in the perceptual space. When looking only at the cards, the consumers' grouping and attribute generation focused in great detail on the nutritional information. The calorie content also became a much more important factor in their perceptions of the bars. Interestingly for the product category analysed, messages about “satiating” and “meal replacement” effects were clearly distinct and negatively correlated in the consumers' minds. © 2014 Elsevier Ltd. All rights reserved.

1. Introduction In recent times, long working hours, changes in lifestyle and changes in family units have brought about changes in eating habits, often altering traditional meal times and the number of meals a day (Mestdag, 2005; Poulain, 2002) leading to fewer real meals being taken, while the number of informal, rapid snacks is increasing. As a result, there has been enormous growth in the development of foods to eat in these situations, such as snack bars (sometimes claimed to be “satiating”) or bars sold as meal replacements (“two bars replace a meal” or similar messages). Consumers find that the information on the

⁎ Corresponding author at: Nofima AS, P.O. Box 210, 1431 Ås, Norway. Tel.: + 47 45426026; fax: +47 64943314. E-mail addresses: paula.varela.tomasco@nofima.no, [email protected] (P. Varela).

http://dx.doi.org/10.1016/j.foodres.2014.07.028 0963-9969/© 2014 Elsevier Ltd. All rights reserved.

packaging of these foods is very diverse and they probably do not fully comprehend it, both because of the variety of messages and because they lack the necessary knowledge. Some of them do not have enough time and want something they can eat quickly that will stop them from feeling hungry for a certain length of time, while others are overweight and think that eating a bar or two instead of a meal will help with weight control. However, hunger is not a single motivation but an interaction of several distinct mental and physical processes, and is therefore influenced by numerous external factors. A recent study highlighted specific potential benefits of satiating food. These include providing appetite control strategies for consumers in general and for those who are highly responsive to food cues; offering pleasure and satisfaction associated with low-energy/healthier versions of foods without feeling “deprived”; reducing dysphoric moods associated with hunger, especially during energy restriction; and improved compliance with healthy eating or weight-management efforts (Hetherington et al., 2013).

M. Miraballes et al. / Food Research International 64 (2014) 456–464

Because of this link between satiety and food intake, satiety is increasingly seen as a source of added value for the agribusiness industry, which has developed satiety-based weight-control products. Formulating products to meet these changing calorie demands and consumer expectations of satiation and satiety (sensations of fullness within meals and between meals, respectively) is a complex process that will continue to be a focus of attention for scientists and food manufacturers (Fiszman & Varela, 2013). The information shown on the packaging is known to have a considerable effect on the consumers' perception of the food (Becker, van Rompay, Schifferstein, & Galetzka, 2011; Schifferstein, Fenko, Desmet, Labbe, & Martin, 2013). The manufacturers use the packaging to provide information and to highlight certain aspects of the product's

457

formulation and certain benefits associated with it (Carrillo, Varela, & Fiszman, 2012a). These messages can be verbal or non-verbal (colours, pictures, lettering size and type) (Carrillo, Varela, & Fiszman, 2012b). They can affect how the consumers perceive the product in different ways, giving rise to different expectations concerning its potential consumption. These expectations can be hedonic, related to sensory attributes, concerning quality, etc. It is quite common for the packaging of products such as snack bars to include statements concerning their satiating effects (“feel less hungry”, “prolonged satiating effect”, etc.) that could have a positive effect on appetite control by generating specific expectations (Brunstrom, 2011; Chambers & Swanson, 2012; Fay, Hinton, Rogers, & Brunstrom, 2011). Some studies have reported examples of these effects, such as heightening the satiating effect of

Table 1 Information provided on the sample packaging. Sample Denomination

Claims

Nutritional informationa

A

Milk chocolate coated bar with magnesium and vitamin E

EV: 341; P: 26; CH: 33.7; SU: 22.4; TF: 11.3; SF: 5.8; γLA: 1; DF: 3.5; Na: 190; M: magnesium; and V: E.

B

Crunchy bar bathed in milk chocolate

This cocoa-flavour bar bathed in milk chocolate makes an appetizing snack. Enriched with magnesium and vitamin E. The calories provided by the carbohydrate, protein and fat content respectively account for 40%, 30% and 30% of the total bar calories. Crunchy. 2 bars replace a meal. Meal replacement for weight control.

C

Meal replacement chocolate bars

2 bars replace 1 meal. With chitosan, fibre, 12 vitamins and minerals.

D

Cereals and black chocolate bar

E

Meal replacement bars for weight control. Chocolate flavour

The ingredients help you meet all your daily vitamin needs while controlling your calorie intake. 97 Kcal per bar. Ideal for taking care of your figure between meals. The low glycaemic index of these bars means that carbohydrates are absorbed more slowly and gradually, helping to dampen your appetite for longer. Feeling less hungry at mealtimes helps you keep your weight down. A treat that staves off hunger for longer between meals. Meal replacement for weight control 1 bar = 1 meal. Helps regulate blood sugar levels. For a safe, effective and pleasant low-calorie diet.

F

Milk chocolate bar with chocolate chips (6%)

G

Multivitamin bars bathed in black chocolate (74%) Meal replacement for weight control

H

I

Chocolate coated satiating bars

Meal replacement bars. 2 bars = 1 meal. A safe, effective way to lose weight. Designed to provide all the necessary nutrients and a controlled calorie content for one complete, balanced meal. With protein-rich foods like these bars you feel fuller for longer, helping you to control your weight. Multivitamin biscuits. This product should not be used as a meal replacement in a balanced diet. Control your weight easily, quickly and safely. 2 bars = 1 meal. New recipe, tastier and more effective. Only 151 Kcal per bar. With 12 vitamins and minerals. Rich in iron. This product complies with the Spanish and European regulations concerning low-calorie weight-loss diets. Satiating bars. Crunchy. With glucomannan and L-Carnitine. Double action: satiating and fat-burning. Keep your appetite under control with this delicious snack. Vitamin enriched. Made with glucomannan, a vegetable fibre that swells up to 100 times its volume in contact with water, and with L-carnitine to encourage and accelerate the burning of accumulated fat. 2 bars = 1 meal.

J

Chocolate flavoured meal replacement for weight control

K

Chocolate flavoured high protein 17 g pure protein. This diet is an excellent, lasting way to bar with sweeteners, 20% lose and control weight successfully. 1 g sugar, 2 g net chocolate, no added sugars carbs per portion. This product can be used at all stages of the Atkins diet. If you are trying to achieve or keep to your ideal weight, these bars are a delicious and nutritious way to achieve your aim.

EV: 375; P: 24; CH: 45; SU: 27; PO: 12; TF: 12; SF: 6; DF: 5; Na: 300; M: calcium, phosphorus, iron, zinc, copper, iodine, selenium, magnesium, manganese, potassium; V: A, D, E, C, thiamine, riboflavin, niacin, B6, folic acid, B12, biotin, and pantothenic acid. EV: 382; P: 26; CH: 38; TF: 12; DF: 9; Na: 212; M: potassium, calcium, magnesium, phosphorus, iron, zinc, copper, selenium, iodine, manganese; V: A, D, E, C, B1, B2, B12, niacin, folic acid, and pantothenic acid. EV: 486; P: 6; CH: 58; SU: 38; TF: 24; SF: 14; DF: 7; Na: 200; V: A, D, E, B1, B2, B3, B6, B9-folic acid, B12, biotin, and pantothenic acid.

EV: 290; P: 27.8; CH: 21.6; SU: 14.4; PO: 6; FR: 5.4; S: 6.5; L: 1.1; TF: 11.3; SF: 4.9; MUF: 3.9; PUF: 2.5; LA: 1.8; αLA: 10; DF: 18.4; Na: 253; M: calcium, phosphorus, potassium, iron, zinc, copper, iodine, selenium, magnesium, manganese; V: A, D3, E, C, B1, B2, niacin, B6, folic acid, B12, biotin, and pantothenic acid. EV: 368; P: 25; CH: 44; SU: 29; PO: 14; TF: 12; SF: 6; DF: 3; Na: 300; M: calcium, phosphorus, iron, zinc, copper, iodine, selenium, magnesium, manganese, potassium; V: A, D, E, C, B1, B2, B3, B6, B9-folic acid, B12, biotin, and pantothenic acid. EV: 472; P: 6; CH: 58; SU: 38; TF: 24; SF: 14; DF: 7; Na: 200; V: A, D, E, B1, B2, B3, B6, B9, B12, biotin, and pantothenic acid. EV: 378.4; P: 24; CH: 43.7; SU: 29; PO: 9; TF: 12; SF: 5.5; UF: 4.8; MUF: 3.2; PUF: 1.6; LA: 1.6; DF: 7; FOS: 3.7; Na: 300; M: calcium, phosphorus, iron, magnesium, zinc, iodine, potassium, copper, manganese, selenium; V: A, D, E, C, B1, B2, niacin, B6, folic acid, B12, biotin, and pantothenic acid. EV: 420; P: 27.5; CH: 42.5; TF: 12.5; DF: 10; Na: 92.5; M: potassium, calcium, magnesium, phosphorus, iron, zinc, copper, selenium, iodine, manganese; V: A, D, E, C, B1, B2, B12, niacin, folic acid, and pantothenic acid.

EV: 402; P: 28; CH: 41; SU: 23.5; TF: 12.5; SF: 8.6; LA: 1.2; DF: 6.5; Na: 200; M: calcium, magnesium, phosphorus, iron, iodine, zinc, potassium, manganese, copper, selenium; V: A, D, E, C, B1, B2, B3, B6, B12, biotin, pantothenic acid, and folic acid. EV: 400; P: 29; CH: 30; SU: 1.6; PO: 27; G: 18; MA: 8.5; ST: 1.5; TF: 21; SF: 12; DF: 11; and Na: 220.

a EV: energy value expressed as kcal/100 g. P: proteins; CH: total carbohydrates; SU: sugars; PO: polyols; G: glycerin; MA: maltitol; FR: fructose; S: sucrose; L: lactose; ST: starch; TF: total fat; SF: saturated fat; UF: unsaturated fat; MUF: monounsaturated fat; PUF: polyunsaturated fat; LA: linoleic acid; αLA: α-linolenic acid; γLA: γ-linolenic acid; DF: dietary fibre; FOS: fructooligosaccharides; Na: sodium; M: minerals; and V: vitamins. All the contents expressed as g/100 g of product; αLA, γLA y Na expressed as mg/100 g of product.

458

M. Miraballes et al. / Food Research International 64 (2014) 456–464

fruit smoothies (Fay et al., 2011). In addition, satiety-related claims have generated much debate in recent years, from both academic and legislative perspectives. Concerns have been raised that satiety claims might be over-interpreted to imply that products deliver weight control or weight loss benefits (Bilman, Kleef, Mela, Hulshof, & van Trijp, 2012). The effect of the packaging design elements on the sensory perception of a food has been amply investigated. Becker et al. (2011) found that different images and different colour tones on lemon yoghurt packs altered the consumers' perception of the product and their expectations concerning its price. Varela, Ares, Giménez, and Gámbaro (2010) studied the influence of brand and packaging on the real and expected acceptability of orange-flavour beverages prepared from a powdered product and found that in certain cases these factors influence the expected quality or acceptance to a very large extent. Projective Mapping is a relatively new option for studying external factors such as these related to perceptions of food products (Varela & Ares, 2012). This method was developed by Risvik, McEwan, Colwill, Rogers, and Lyon (1994) and reintroduced by Pagès (2005) under the name of Napping®. Its main advantage is that it is useful for studying differences and resemblances between products from a global viewpoint, and as it is holistic it also obtains more spontaneous responses than other techniques (Guerrero et al., 2010). In this method, consumers are asked to arrange a group of samples on a plane, following their own criteria, with the only requirement that closeness between the samples indicates similarities and distance indicates differences (Risvik, McEwan, & Rødbotten, 1997). Napping® is usually associated with ultra-flash profiling (UFP) (Pagès, 2003; Perrin et al., 2008) as the two techniques show a high degree of complementarity: once the samples have been arranged on the plane, UFP provides the descriptions to explain the differences encountered. The aim of this study was to investigate the effect of extrinsic factors on consumers' perceptions of satiating and meal replacement bars in two different scenarios: examining the complete packaging (the box), or only the product denomination, messages and nutritional information taken from the box, as a way of obtaining greater insight into the consumers' knowledge and interpretation of the concept “satiating” and related terms. 2. Materials and methods 2.1. Samples Eleven samples of commercial chocolate flavoured bars labelled as “meal replacement” and/or “with satiating effect” were employed. The samples were bought in supermarkets and parapharmacies in the city of Valencia (Spain). The packaging information is shown in Table 1. 2.2. Consumers The study was conducted in the city of Valencia (Spain) with a convenience sample of 60 women aged between 19 and 68 years (mean: 40.0, standard deviation: 12.1) recruited ad hoc. The reason for selecting only women was that the product under study is mainly consumed by this section of the population. The consumers were divided into two sociodemographically homogeneous groups of 30 women each. One Napping® session was held with each group, with a different scenario in each session. Convenience samples are usually used in qualitative studies that aim for an approximation to a research subject. They involve recruiting available participants who meet specific criteria (Kinnear & Taylor, 1999). Also, in the present case the intention was not to obtain a reading of consumer liking or market positioning, but to elicit a description of a group of products in terms of sensory and nutritional perceptions. The criterion for selecting the participants was their interest in participating in the study. At the recruitment stage, no information about the specific aim of the study was provided.

2.3. Napping® The exercise was completed in a session lasting approximately 20–30 min. The consumers were first introduced to the method through an example using geometrical figures with different colours. They were told that placing the samples close together on the sheet of paper provided (A3, 60 × 40 cm) meant that they were similar, while placing the samples apart meant that they were different (Carrillo et al., 2012b). At no point were they told what product they would be assessing, or any of its characteristics, in order to avoid bias. Each consumer was then given the 11 samples (see the description of the two sessions at Section 2.4 below), all together, and was asked to look at them and arrange them on a sheet of paper according to their differences and similarities, following her own criteria (Pagès, 2005). Each sample was marked with a three-digit code and the consumers were provided with self-adhesive tags with the sample codes to arrange on the sheet of paper using all the available space. They were then asked to write comments beside each code or group of codes to say why they had placed the samples where they had (Perrin et al., 2008). 2.4. Procedure Two Napping® sessions (different evaluation scenarios, as described below) were held with the two groups of participants, employing a different group of consumers (n = 30) in each session. 2.4.1. Session 1 The first group of 30 consumers assessed samples consisting of the bars' packaging. The eleven samples were coded with random threefigure numbers. The consumers were provided with a sheet of paper and self-adhesive tags with the sample codes and were asked to arrange the tags on the paper according to the similarities and differences between the products, and then write beside them the characteristics they had used to position the tags or groups of tags in the Napping® task. 2.4.2. Session 2 The cards designed for this task contained only the product denomination, the nutrition information panel and the health and nutritional claims on the packaging of each bar (11 samples, see the example in Fig. 1). All the cards were given to the participants to evaluate, explaining that the information on the cards corresponded to the bar package information. The participants had to flick through the cards and place them on the A3 sheet of paper according to the similarities and differences they perceived in the potential products corresponding to the information on the cards, using their own criteria. They then had to write alongside the characteristics they had used to position the cards or groups of cards on the paper in the Napping® task. The 11 cards were all printed in black and white, with the same font size, typography and format, so that no design features (position on the package, size, font colour, etc.) would influence the participants. Each card was coded with a random three-digit number. 2.5. Data analyses All the words provided by the participants in the ultra-flash profiling step following the projective mapping task were qualitatively analysed for each session separately. The terms elicited to describe a sample or group of samples were grouped together, using synonymous and derived words, by triangulation (consensus between three researchers after separate processing). The frequency of mention was determined by counting the number of mentions of the same term in each session. Frequency of mention of synonyms and repeated sensory attributes were combined and considered as one variable in the data analysis

M. Miraballes et al. / Food Research International 64 (2014) 456–464

459

Table 2 Classification of the terms used to describe the samples at the two sessions, elicited by the ultra-flash profile technique. Category

Attribute

Nutritional information

1 bar = 1 meal 2 bars = 1 meal Balanced Fat burner For weight control Glycaemia control Healthy High energy content With nutritional information/ similar nutritional values Medium energy content Not for meal replacement Satiating/for appetite control Specific diet Meal replacement Without nutritional information/ insufficient information/little information about fat and sugar content Enriched with vitamins and minerals High fibre content High protein content High saturated fat content High sodium content High sugar content High fat content

Ingredients/composition

Fig. 1. Example of a card used in session 2, showing the information displayed on the bar packaging.

(i.e.: dull, mild, bland would be entered as dull/mild/bland and its frequency of mention would be the sum of the frequencies of each word). Only terms that had been mentioned at least three times were used for the analysis and a frequency table was built for each session. A chi square test was used to test significant differences between sessions. The product positioning (X, Y coordinates) were measured in centimetres considering the bottom left corner of the paper sheet as origin of the coordinates (0,0), and the attributes were listed together with their frequencies of mention across the consumer panel. MFA was performed with XLStat system software (version 2009.4.03, Addinsoft™). It was performed on the X and Y coordinate values for the samples on each of the participants' individual maps, covering both scenarios. The data table consisted of 11 rows representing the samples and 30 X and Y coordinates for each scenario (participants) as suggested by Pagès (2005). The frequency table containing the terms generated by the consumers and their frequencies of mention was considered a set of supplementary variables and did not contribute to the construction of the MFA factors. MFA allows the tables, the variables, the principal axes of the analyses, and the individuals to be visualized in a two or three dimensional space (Escofier & Pagès, 1984). The selection of the number of relevant factors was based on the scree plot of the eigenvalues. Hierarchical Cluster Analysis (HCA), using Euclidean distances and Ward's aggregation criterion, was carried out in order to identify samples with similar characteristics within each scenario. 3. Results and discussion

Branding/uses/attitudes

Sensory-related expectations

Packaging design

L-carnitine Low fat content Low fibre content Low protein content Low sodium content Medium fibre content Medium protein content Medium sodium content Only Vitamin E/only Mg With little sugar With natural additives With sugar Credible I like it less I like it more Snack Practical/a good size Cake Black or milk chocolate Compact appearance Biscuit/coated biscuit Crunchy Dark chocolate Not a biscuit Thicker bar Milk chocolate Mousse inside Chocolate Similar texture Thin bar Little or no chocolate Big packaging Similar packaging and colours White packaging

Packaging Cards 7

8

5 26 3 3 14 3

5 36 13

23 29 3 76 13

14 24 17 4 80 13

4 7 19 3 13 11 12 6 11 7 16 10 5 5 3 18 3 20 15 3 7 4 24 6 3 4 4 18 9 53 6 6 34 6 13 3 5

14

6 8 4

7 39

3 3 6 3

Note: Chi square tests highlighted significant differences between the two experimental sessions in all the categories.

3.1. Associations generated through ultra flash profiling The terms mentioned most often by the consumers in the two Napping® sessions were classified into five categories: “Nutritional information”, “Ingredients/Composition”, “Branding/Uses”, “Sensoryrelated expectations”, and “Packaging design” (Table 2). Chi square tests highlighted significant differences between the two experimental sessions in all these categories.

3.1.1. Nutritional information The most frequent terms within this category were related to replacement/not for meal replacement, followed by satiety/appetite control. Those terms were cited with similar levels of frequency in both sessions, showing their importance for this category of products. However, the meal replacement “power” (number of bars needed to replace a meal) was mentioned more in the cards scenario (session 2)

460

M. Miraballes et al. / Food Research International 64 (2014) 456–464

than in the more realistic scenario of looking at the packaging (session 1). Other nutritional properties such as calorie content were only mentioned when looking at the cards. This is somewhat surprising for this category of products, since calories would be a logical property to look at, but at the same time it highlights the fact that consumers relied more on the claims than on nutritional data, like calorie content, for their categorization. According to Bilman et al. (2012), expert opinions seem to diverge on whether satiety claims on products play a role as an adjunct to weight control (de Graaf, 2011; Mela, 2011), or are interpreted as a guarantee of weight control and even weight loss success (Bellisle & Tremblay, 2011; Booth & Nouwen, 2010, 2011; Smeets & van der Laan, 2011), carrying the risk that consumers may abdicate their own responsibility for a broader repertoire of voluntary behaviour. Another aspect of this could be related to the test; nutritional cards present the information somehow “artificially” to the consumer, making them focus in the nutritional information more than they would do in real settings, which might also happen in other studies of label understanding or attention to, e.g. via conjoint analysis or eye tracking. In this sense, Varela, Antunez, Silva Cadena, Giménez, and Ares (submitted for publication) have found studying attentional capture in real packs via mobile eye tracking that when information is presented in a single plane using product cards or a computer monitor, attentional capture is modified and the salience of the displayed information increased, which may lead to an overestimation of the influence of certain parameters on food choice. 3.1.2. Ingredients/composition Regarding the “Ingredients” category, only few references were made to specific compounds or ingredients when evaluating the packaging (references only to sugar and in general to the enrichment with vitamins and minerals). When looking only at the cards, however, associations were made with specific compounds (like vitamins, minerals, proteins or fibre) and their quantities (low-, medium-, high-fibre, protein, sodium). It seemed that when the consumers had the information, with the same font size and colour of the rest, in sight (same plane and not in the back of the packaging box where they would or would not read it, and normally in a smaller font size) they tended to classify it, though no specific knowledge about the functional relevance of each ingredient was shown. Individual differences in health consciousness may be an important moderating factor in the impact of health-related labels on perceptions (Vadiveloo, Morwitz, & Chandon, 2013). In the present case a long list of vitamins and minerals might have had some influence and have been interpreted as indicating a healthy or complete food product. A study by Finkelstein and Fishbach (2010) reported higher hunger ratings after people consumed a small snack food with a label that emphasized the healthfulness rather than the tastiness of a snack, suggesting that a very healthy snack is not considered filling. This distinctive behaviour regarding the two scenarios agrees with the comments on calorie content, again stressing the low importance attached to nutritional labelling when consumers are faced with interpreting or choosing the complete box in real life situations. Similar findings were highlighted by Carrillo et al. (2012b) for a different type of products (enriched and reduced-calorie biscuits): the consumers would focus more on claims than on the objective nutritional information or ingredient list. The exception in the present work was the sugar content (“with sugar”/“with little sugar” were emphasized) and the reference to “with natural additives”, mentioned when looking at the box. This finding is in agreement with those of Childs, Thompson, Lillard, Berry, and Drake (2008), who held consumer focus groups about different meal replacement products and stated that health claims were generally met with scepticism by most consumers, but that there were some claims which panellists actively looked for on packages, such as “natural” and “low sugar”. Through conjoint analysis these authors found that the top three factors influencing the purchase of meal replacement products (bars and beverages) were nutritional contents,

flavour and price. However, protein type and content and other nutritional facts were actively introduced into the focus group task by the leader, since the perception of these nutrients was the principal objective of that study, so it was logical for nutritional contents to appear as an important factor for purchasing. The results of Childs et al. (2008) might bear a greater resemblance to what happened in the cards scenario in the present study and might not be representative of what would happen in a real context, when the choice would be based on the box information, as the consumers in the box situation in the present study did not actually focus on the detailed nutritional content of the bars. 3.1.3. Sensory-related expectations The “sensory” expectations category contained several associations, mainly made when looking at the packaging. However, some sensory terms were also elicited in the cards exercise, very probably determined by the product denomination: for example “Chocolate-flavoured bars, one meal replacement, for weight management” would lead consumers to focus on the chocolate flavour. If the image of the product was included on the front of the box it also elicited some sensory associations related to texture and appearance (crunchy, mousse, compact, thick or thin) or taste (mainly chocolate-related). 3.1.4. Branding/uses/attitudes This is a minor category in terms of the frequency of terms elicited. They were mainly related to liking (I like it more, I like it less) or to use as a snack and, as expected, more terms were generated in session 1 (when looking at the packaging). It is important to highlight that “credible” and “practical” were positive and that no negative terms were obtained, such as “not true” and “excessive marketing” as reported by Carrillo et al. (2012a), who found that excessive information or a high number of claims on the pack had a negative influence on the perception and expected acceptance of low calorie/functional biscuits. Williams (2005) pointed out that consumers prefer split claims, with a succinct statement on the front of the package and more detailed information provided elsewhere on the package. 3.1.5. Packaging design This is another minor category among the elicited terms. It refers to the colours and sizes of the packaging evaluated in session 1, so evidently session 2 did not generate any terms in this category. 3.2. Session 1. Perception of the box The HCA only highlighted three groups in the box scenario session: cluster 1 (C1, samples C, J), cluster 2 (C2, samples B, E, F, H), and cluster 3 (C3, samples A, D, I, G, K). These three groups also made sense in the MFA product map (Fig. 2a), as the samples were clearly sorted in the perceptual space in the same way. In this session, most of the words or phrases used to describe the samples or groups of samples fell into the “sensory-related expectations” category, followed by “ingredients” and “nutritional information”, with a smaller percentage in the “branding/uses”, and “packaging design” categories. The three first factors of the MFA accounted for 61.62% of the total variability of the sample set in session 1. The first factor (30.81% of explained variance) separated the samples into those declared to be satiating (with no mention of meal replacement), in the positive values (samples D, G, I) and the samples that claimed to be a meal replacement (B, E, F, H, J, C), in the negative values, while samples A and K, which did not state either of those concepts on the box, were plotted mid-way on the map. The second factor (17.98%) mainly separated samples J and C, the ones with least nutritional information and claims, from the rest, but sample K was also separated towards the upper (but in this case right-hand) quadrant of the perceptual space, probably because of it was targeted towards a special diet (“Enerzone”) that may not have been familiar to the consumers.

M. Miraballes et al. / Food Research International 64 (2014) 456–464

Fig. 2. Perceptual space determined by the first two factors of the MFA in session 1; (a) sample plot; (b) attribute plot. Ellipses on the sample plot show the grouping obtained via HCA.

The important fact is that the main factor separating the samples was whether they were stated to be satiating or for meal replacement. These two aspects were negatively correlated, suggesting that they were differentiated in the consumers' minds and are not interchangeable as could be expected. It is interesting, for example, that the sample G box, which states that this bar is “not for meal replacement”, was placed close to the satiating products. This result is not in accordance with the fact that meal replacement products intended for weight control would generally be designed to be satiating (Rothacker & Watemberg, 2004). Bilman et al. (2012), on the other hand, found that consumers interpreted satiety claims quite accurately, staying close to their literal meaning. In a recent study, Fiszman, Varela, Díaz, Linares, and Garrido (2014) stated that in the consumers' minds, the meaning attached to “satiety” is more in line with the dictionary meaning – (1) the quality or state of being fed or gratified to or beyond capacity: surfeit, fullness; and (2) the revulsion or disgust caused by overindulgence or excess (Merriam-Webster Dictionary, 2011) – than with the academic meaning reported by de Graaf (2011), namely a subjective feeling of a reduction in the motivation to eat. Further to this, the fact of satiating and meal replacement being negatively correlated could

461

also suggest that consumers attached an extra meaning to meal replacement, as being nutritionally balanced. In terms of sample grouping, as highlighted by the HCA, cluster C1 (samples C and J) was described in terms of “without nutritional information”, “2 bars = 1 meal”, “looks like chocolate only”, “dark” and “not sure if it is dark or milk chocolate”, as well as “white packaging” and “big packaging”. These two samples were those that made almost no claims, mainly displaying only their use as meal replacements (2 bars = 1 meal). These products did not have pictures on the box, so consumers were less sure about what to expect in terms of their sensory aspects. As a result, fewer sensory attributes were associated with these samples and they were placed apart. Cluster C2 (samples B, E, F, H) was described as: “Meal replacement”, “for weight control”, “similar nutritional values”, “Mg enriched”, “with nutritional information”, “with chocolate”, “practical”, “a good size”, “similar texture” and “cake”. All the products in this cluster bore the words “meal replacement” in large letters on the front of the packaging, the packaging were of similar sizes and the front of the box also showed a bar cut in half (milk chocolate colour), showing the interior texture. This last fact probably influenced the generation of more attributes related to their sensory characteristics. Carrillo et al. (2012b) found similar results when studying the extrinsic properties of reducedcalorie biscuits; even when not tasting them but only looking at the box, consumers already thought about many sensory attributes they might expect when consuming the product. The messages on the packaging of the four samples in cluster C2 highlighted their use for weight control/low calorie diets. However, other samples grouped elsewhere also displayed this benefit (I, D or K). This might point again to weight control being mainly associated with meal replacement rather than with satiation, in accordance with the findings of Bilman et al. (2012), who discussed whether consumers do not over-interpret satiety claims as related to or implying weight loss. However, consumers would perceive meal replacement products as being more related to weight loss. It has to be pointed out, nevertheless, that over-interpretation of this kind of claims would also be related to the particular characteristics of the consumer (e.g. overweight vs. normal weight, men vs. women), a fact that has not been studied in the present work. Also, different kinds of consumers might focus differently on different claims, as highlighted by Childs et al. (2008) in their work with whey and soy protein meal replacement food items, who found that exercisers viewed muscle-building claims as more important than non-exercisers, while non-exercisers perceived heart health, calcium and vitamin/mineral claims as more important than exercisers. Cluster C3 (samples A, D, I, G, K) was associated with descriptions such as “satiating”, “for appetite control”, “not for meal replacement”, “with vitamins and minerals”, “thin bar”, “coated biscuit”, “snack” and “healthy”. Looking into factors 1 and 3 of the MFA (30.81% of the total variability, plot not shown), it was found that these two dimensions separated two subgroups, one made up of samples A and K (products for specific diets such as Atkins and Enerzone) and the other comprising samples I, G and D, associated with “satiating/for appetite control”, “cake”, “thin bar” and “enriched with vitamins and minerals”.

3.3. Session 2. Perception of the cards Regarding the generation of words through UFP in session 2, this differed from the results of session 1, where the packaging was shown, in that most of the words or phrases used to describe the samples or groups of samples fell into the “ingredients” and “nutritional information” categories, which is logical, though a smaller percentage related to “branding/uses” and “sensory-related expectations”. Evidently, no mention of packaging design was found. HCA highlighted 3 groups of samples: cluster C4 (samples E and K), cluster C5 (samples A, D, G) and cluster C6 (samples B, C, F, H, I, J), which

462

M. Miraballes et al. / Food Research International 64 (2014) 456–464

were very much in line with the positioning in the MFA perceptual space defined by the first two factors (Fig. 3a). According to the MFA (F1 vs F2) (Fig. 3b): C4 was associated with “high energy content”, “specific diets”, “1 bar = 1 meal”, “high protein content”, “high fibre content” and “high fat content”. C5 was associated with terms such as “not for meal replacement”, “low energy (low fat/low sugar) content”, “low protein content”, “no vitamins or minerals”, “only vitamin E or Mg”, “balanced”, “biscuits”, “dark chocolate” and “snack”. C6 was described as “meal replacement”, “2 bars = 1 meal”, “medium energy content”, “medium protein content”, “little information on sugar content”, “fat”, “saturated fat” and “little chocolate”.

The vitamin and mineral contents were also important for the consumers, who grouped the samples on this basis: cluster C5 is made up of two samples with no mineral content information (samples D and G) and one that only mentioned Vitamin E and Mg (sample A), whereas all the other samples had high vitamin and mineral contents.

3.4. Comparison of the two sessions

Clearly, the consumers mainly differentiated between the samples on the basis of the differences in nutritional facts, paying most attention to the energy, protein and fat contents. The results are consistent with the nutritional information on these contents shown on the cards (the same as on the packaging — see Table 1). Also, since there was no great difference in carbohydrate content between the different samples, the consumers were coherent in their references to this fact.

The superimposed representation of the samples in the MFA made it possible to observe the proximity between the two evaluation instances for each evaluated sample (Fig. 4). The MFA was performed on the X–Y coordinate values of the samples. The Rv coefficient obtained was quite low (0.68), showing that the samples were perceived differently in the two sessions. The Rv coefficient is a measure of the similarity between two factorial configurations. It would be 0 if the configurations were uncorrelated and 1 if the configurations were homothetic or correlated completely. This coefficient depends on the relative position of the points in the configuration and is independent of rotation and translation (Robert & Escoufier, 1976). The samples A, E, G, I and K were the ones that presented the bigger distances between the two sessions. It is difficult to discern what the reason for these differences was in particular since the visual cues in

Fig. 3. Perceptual space determined by the first two factors of the MFA in session 2; (a) sample plot; (b) attribute plot. Ellipses on the sample plot show the grouping obtained via HCA.

Fig. 4. Superimposed representation of the samples in the MFA of the two sessions; (a) sample plot; (b) attribute plot. Each sample is represented using two points corresponding to each evaluation instance: packaging and cards. The consensus representation that takes into account both evaluation instances is also represented.

M. Miraballes et al. / Food Research International 64 (2014) 456–464

the packaging were various: kind and size of typography, colours, product photographs, etc. Similar results were reported by Carrillo et al. (2012a), who found that for enriched biscuits, different products could have different effects on the extrinsic cues influencing consumer perception, even within the same category, as in the present case. The chi-squared test that compared the two sessions to detect whether there were significant differences between the descriptive terms or phrases used by the different groups showed that the results differed (Table 2). This indicates that the more natural evaluation scenario where the participants had to observe, compare and assess the whole packaging, which is closer to what they would do when buying in a shop (session 1), was indeed different from the other evaluation condition where they were directed to focus on the information they read on the card (the composition and nutritional characteristics of the product) in order to generate their perceptions. As already discussed above, the sensory terms that emerged in session 1 (with the packaging) showed considerable variety as they reflected the sensory expectations that the pictures, shapes, colours etc. aroused among the consumers. In general, the big difference between the two sessions was that in grouping and describing the samples, the consumers in session 1 (with the packaging) paid more attention to the pictures and claims and placed less emphasis on the nutritional information and ingredients. The words elicited in the two sessions were also compared through MFA. The Rv value obtained in this case was also low (0.7), and the overlaid attribute plot (Fig. 4) shows that only some of the attributes were correlated between the two scenarios, namely those related to the product denomination (“with chocolate”, “snack”) or those that appeared in the claims or messages highlighted on the packaging such as “enriched with vitamins and minerals”, “satiating”, “meal replacement”, “2 bars = 1 meal”, or “low sugar”. Detailed nutritional aspects of the products, like specific levels of nutrients, specific mineral contents, etc., were not used for sample categorization. The results showed that the nutritional information could potentially be interpreted correctly by the consumers, but when it was accompanied by all the rest of the information (both verbal and non-verbal) on the packaging, its effect was diluted and it was used to a lesser extent in forming a perception of the product. It has to be highlighted that in session 2 (experimental cards) the consumers did not receive any different information compared with session 1 (all this information was on the packaging), but it was given without the distracting (or attractive) effect of other packaging design features. The use of experimental cards in consumer studies is quite common when performing informed tastings or conjoint analysis studies. Consequently it is pertinent to question the extent to which these approaches are valid, since presenting the information in a situation far removed from the “ecological” context could be misleading: rather than just looking at the information, as they would in a natural choice situation such as a supermarket, the consumers would perform the test in much greater depth. Carrillo et al. (2012b) stated that the packaging is really a modulator of the consumers' global perception of the product, which sensory and non-sensory cues interact to elicit. In the particular case of reformulated products with health or nutritional claims, depending on the particular product, parameters like brand, category, familiarity with the product, familiarity with the claim, or sensory profile influenced the assessment, acceptance and perceived healthiness of the product in different ways (Carrillo et al., 2012a). 4. Conclusions The information displayed on the bars' packaging, including messages or claims, had a strong influence on the consumers' perception of the product. This information was observed to have a clear influence on their expectations of the product and to constitute a driver for categorizing the different samples in their minds. An interesting finding of this study was that the effects of “satiating” and “meal replacement” claims were clearly distinct and negatively

463

correlated, implying that consumers do not consider that a meal replacement food product should be satiating, or at least do not demonstrate any evident linking. It would be interesting to test these results in further categories of products. Even though the bars used in the present study are a product category that actively targets nutritional and health issues and nutritionally concerned consumers, the consumers did not consider the detailed nutritional information when evaluating the packaging, whereas they did pay careful attention to it in their categorization when faced with the same information but without the somewhat distracting elements of the pack design (colours, pictures, font size). Acknowledgements This work has received support from the Spanish Ministry of the Economy and Competitiveness (MINECO) through Project AGL201236753-C02-01. The authors wish to thank Mary Georgina Hardinge for assistance with the English manuscript. References Becker, L., van Rompay, T. J. L., Schifferstein, H. N. J., & Galetzka, M. (2011). Tough package, strong taste: The influence of packaging design on taste impressions and product evaluations. Food Quality and Preference, 22(1), 17–23. http://dx.doi.org/10.1016/ j.foodqual.2010.06.007. Bellisle, F., & Tremblay, A. (2011). Satiety and body weight control. Promise and compromise. Comment on ‘Satiety. No way to slim’. Appetite, 57, 769–771. Bilman, E. M., Kleef, E. Van, Mela, D. J., Hulshof, T., & van Trijp, H. C. M. (2012). Consumer understanding, interpretation and perceived levels of personal responsibility in relation to satiety-related claims. Appetite, 59(3), 912–920. http://dx.doi.org/ 10.1016/j.appet.2012.07.010. Booth, D. A., & Nouwen, A. (2010). Satiety. No way to slim. [Article]. Appetite, 55, 718–721. http://dx.doi.org/10.1016/j.appet.2010.08.009. Booth, D. A., & Nouwen, A. (2011). Weight is controlled by eating patterns, not by foods or drugs. Reply to comments on “Satiety. No way to slim”. Appetite, 57, 784–790. Brunstrom, J. M. (2011). The control of meal size in human subjects: A role for expected satiety, expected satiation and premeal planning. The Proceedings of the Nutrition Society, 70(2), 155–161 (Retrieved from http://europepmc.org/abstract/MED/ 21275082). Carrillo, E., Varela, P., & Fiszman, S. (2012a). Effects of food package information and sensory characteristics on the perception of healthiness and the acceptability of enriched biscuits. Food Research International, 48(1), 209–216. http://dx.doi.org/10. 1016/j.foodres.2012.03.016. Carrillo, E., Varela, P., & Fiszman, S. (2012b). Packaging information as a modulator of consumers' perception of enriched and reduced-calorie biscuits in tasting and non-tasting tests. Food Quality and Preference, 25(2), 105–115. http://dx.doi.org/ 10.1016/j.foodqual.2012.02.005. Chambers, J., & Swanson, V. (2012). Stories of weight management: Factors associated with successful and unsuccessful weight maintenance. British Journal of Health Psychology, 17(2), 223–243. http://dx.doi.org/10.1111/j.2044-8287.2011.02030.x. Childs, J., Thompson, J., Lillard, J., Berry, T., & Drake, M. (2008). Consumer perception of whey and soy protein in meal replacement products. Journal of Sensory Studies, 23, 320–339. de Graaf, C. (2011). Trustworthy satiety claims are good for science and society. Comment on ‘Satiety. No way to slim’. Appetite, 57, 778–783. Escofier, B., & Pagès, J. (1984). L'analyse factorielle multiple: Une méthode de comparaison de groupes de variables. Data Analysis and Informatics, III. Fay, S. H., Hinton, E. C., Rogers, P. J., & Brunstrom, J. M. (2011). Product labelling can confer sustained increases in expected and actual satiety. Appetite, 57(2), 557. http://dx.doi.org/10.1016/j.appet.2011.05.069. Finkelstein, S. R., & Fishbach, A. (2010). When healthy food makes you hungry. Journal of Consumer Research, 37, 357–367. Fiszman, S., & Varela, P. (2013). The satiating mechanisms of major food constituents — An aid to rational food design. Trends in Food Science & Technology, 32(1), 43–50. http://dx.doi.org/10.1016/j.tifs.2013.05.006. Fiszman, S., Varela, P., Díaz, P., Linares, M. B., & Garrido, M.D. (2014). What is satiating? Consumer perceptions of satiating foods and expected satiety of protein-based meals. Food Research International, 62, 551–560. Guerrero, L., Claret, A., Verbeke, W., Enderli, G., Zakowska-Biemans, S., Vanhonacker, F., et al. (2010). Perception of traditional food products in six European regions using free word association. Food Quality and Preference, 21(2), 225–233. http://dx.doi. org/10.1016/j.foodqual.2009.06.003. Hetherington, M. M., Cunningham, K., Dye, L., Gibson, E. L., Gregersen, N. T., Halford, J. C. G., et al. (2013). Potential benefits of satiety to the consumer: Scientific considerations. Nutrition Research Reviews, 26(1), 22–38. http://dx.doi.org/10.1017/ S0954422413000012. Kinnear, T., & Taylor, J. (1999). Investigación de Mercados Un enfoque Aplicado (5ta. Edición ). México: Editorial Mc Graw Hill.

464

M. Miraballes et al. / Food Research International 64 (2014) 456–464

Mela, D. J. (2011). Satiety. Let's put claims in the right context. Comment on ‘Satiety. No way to slim’. Appetite, 57, 774–777. Merriam-Webster Dictionary (2011). Merriam-Webster.com. http://www.merriamwebster.com (6 December 2013) Mestdag, I. (2005). Disappearance of the traditional meal: Temporal, social and spatial destructuration. Appetite, 45(1), 62–74. http://dx.doi.org/10.1016/j.appet.2005.03.003. Pagès, J. (2003). Direct collection of sensory distances: Application to the evaluation of ten white wines of the Loire Valley. Sciences Des Aliments, 23(5–6), 679–688. Pagès, J. (2005). Collection and analysis of perceived product inter-distances using multiple factor analysis: Application to the study of 10 white wines from the Loire Valley. Food Quality and Preference, 16(7), 642–649. http://dx.doi.org/10.1016/ j.foodqual.2005.01.006. Perrin, L., Symoneaux, R., Maître, I., Asselin, C., Jourjon, F., & Pagès, J. (2008). Comparison of three sensory methods for use with the Napping® procedure: Case of ten wines from Loire valley. Food Quality and Preference, 19(1), 1–11. http://dx.doi.org/ 10.1016/j.foodqual.2007.06.005. Poulain, J. P. (2002). The contemporary diet in France: “De-structuration” or from commensalism to “vagabond feeding”. Appetite, 39(1), 43–55. http://dx.doi.org/10. 1006/appe.2001.0461. Risvik, E., McEwan, J. A., Colwill, J. S., Rogers, R., & Lyon, D. H. (1994). Projective mapping: A tool for sensory analysis and consumer research. Food Quality and Preference, 5(4), 263–269. http://dx.doi.org/10.1016/0950-3293(94)90051-5. Risvik, E., McEwan, J. A., & Rødbotten, M. (1997). Evaluation of sensory profiling and projective mapping data. Food Quality and Preference, 8(1), 63–71. http://dx.doi.org/ 10.1016/S0950-3293(96)00016-X. Robert, P., & Escoufier, Y. (1976). A unifying tool for linear multivariate statistical methods: The RV-coefficient. Applied Statistics, 25, 257–265.

Rothacker, D.D. Q., & Watemberg, S. (2004). Short-term hunger intensity changes following ingestion of a meal replacement bar for weight control. International Journal of Food Sciences and Nutrition, 55(3), 223–226. http://dx.doi.org/10.1080/ 09637480410001734111. Schifferstein, H. N. J., Fenko, A., Desmet, P.M.A., Labbe, D., & Martin, N. (2013). Influence of package design on the dynamics of multisensory and emotional food experience. Food Quality and Preference, 27(1), 18–25. http://dx.doi.org/10.1016/j.foodqual.2012. 06.003. Smeets, P. A.M., & van der Laan, L. N. (2011). Satiety. Not the problem, nor a solution. Comment on ‘Satiety. No way to slim’. Appetite, 57, 772–773. Vadiveloo, M., Morwitz, V., & Chandon, P. (2013). The interplay of health claims and taste importance on food consumption and self-reported satiety. Appetite, 71, 349–356. Varela, P., Antunez, L., Silva Cadena, R., Giménez, A., & Ares, G. (2014). Attentional capture and importance of package attributes for consumers' perceived similarities and differences among products: A case study with breakfast cereal packages. (submitted for publication). Varela, P., & Ares, G. (2012). Sensory profiling, the blurred line between sensory and consumer science. A review of novel methods for product characterization. Food Research International, 48(2), 893–908. http://dx.doi.org/10.1016/j.foodres.2012.06.037. Varela, P., Ares, G., Giménez, A., & Gámbaro, A. (2010). Influence of brand information on consumers' expectations and liking of powdered drinks in central location tests. Food Quality and Preference, 21(7), 873–880. http://dx.doi.org/10.1016/j.foodqual.2010.05. 012. Williams, P. (2005). Consumer understanding and use of health claims for foods. Nutrition Science and Policy, 256–264.