- Email: [email protected]
Influence of label and location of testing on acceptability of cream cheese varying in fat content
Appetite, 1995, 24, 101-106
Influence of Label and Location of Testing on Acceptability of Cream Cheese Varying in Fat Content
Bi~ATRICE DAILLANT-SPINNLER and SYLVIE ISSANCHOU Laboratoire de Recherches sur les Ar6mes, INRA, Dijon
The acceptability of low-, medium- and high-fat versions of a cream cheese to consumers was measured in different conditions by rating overall distance from ideal. The influence of label (unlabelled vs. commercial package) differed according to the cream cheese fat-level habits of the assessors and to the fat level of the product. In labelled testing, high-fat cream cheese was more highly accepted by high- and medium-fat users and low-fat cream cheese seemed to be further from ideal for high-fat users. The location of testing (home vs. laboratory) affected the rating of only the high-fat product: consumers are more severe on it at home than in the laboratory.
INTRODUCTION Many authors underline the importance of context as an influence on food choice (e.g. K6ster, 1981). This paper deals with the influences on acceptability of a cream cheese by two non-food contextual influences (as defined by Rozin & Tuorila, 1993): labelled fat content and location of testing, in the home and in a laboratory. With respect to effects of information on fat content, Eiser et al. (1984) found that children (13-14 years) rated food high in fat content less pleasant than food low in fat content when information was given, whereas the contrary was observed with control subjects who received no information. K6ster et al. (1987) found that the relative amount consumed of a low-fat product (vegetable rolls) was less when health information was given than when no information was given. Solheim (1992) described for sausages an effect of fat-content information depending on the fat level and/or the sensory difference between low-fat and high-fat versions. Hellemann et al. (1990) reported that high-fat spreads were better liked in labelled testing. Light et al. (1992) reported tendencies (p<10%) for hedonic responses to full-fat and lowfat ice cream to be affected by nutrition information: ice cream with information may have been liked more than ice cream presented without information. Thus information on fat content may increase or decrease acceptability; one explanation for differences in effect could be that the way information on fat content influences acceptability depends on the attitudes of the assessors towards health. Such interaction between consumer attitudes and information on the product was considered by The authors would like to thank P. Schlich for his help in planning the experiment and analysing the data. Address correspondence to: B. Daillant-Spinnler, 1NRA, Laboratoire de Recherches sur les Ar6mes, 17 Rue Sully, BV 1540, 21034 Dijon Cedex, France. 0195-6663/951020101 +05 $08.00/0
© 1995 Academic Press Limited
102
B. DAILLANT-SPINNLERAND S. ISSANCHOU
Aaron et al. (1994): according to these authors, information leads to a shift of hedonic judgments in a direction more consistent with attitudes of the subject. Regarding effects of location of testing, Mailer et al. (1980) reported that military hospital patients eating on the wards expressed more favorable opinions about the food than those eating in the dining room. Meiselman et al. (1988) compared two studies, one in the field and another one in a dining room, with two different groups of soldiers but the same three rations per day. There was a difference of consumption between the groups which these authors attributed to the situational differences. However, although soldiers ate more in the dining room, pleasantness seemed higher in the field (a mean of 7.05 vs. 6"05 in the dining room). In a recent study (Note 1), the items of a meal were rated more pleasant in the laboratory in the first week of the study than in a canteen in weeks 2-5. As it seems that acceptability depends not only on the context but also on the interactions between the context and characteristics of the food or of the consumer, the aim of this work was to pay particular attention to the interaction of the test context with fat level and with consumer habits. The two experiments presented here are part of a larger study whose other results are to be reported separately (Note 2).
METHOD Products
Three types of the same brand of cream cheese (i.e. a dairy product close to cottage cheese but with a smooth texture) were presented to the consumers, with fat contents of 10%, 20% and 40% of dry weight (or 1.5%, 3% and 8% of total weight). Design
Experiment 1 was conducted to study the effect of label on the distance from ideal. The consumers tested the products first blind and then in their commercial package. Only fat content was indicated on the commercial packaging with a color corresponding to each fat content (pink for 10%, blue for 20% and green for 40%). It should be noted that, in France, this color code is the same whatever the brand of cream cheese. Eighty-two consumers participated to this test; 29 reported at the beginning of the study being regular users of low-fat cream cheese, 12 of mediumfat and 30 of high-fat whereas 11 reported no particular cheese habits. In experiment 2, the same products were presented in blind tests in the laboratory and at home. Fifty-two consumers participated in this test; 22 reported being regular users of low-fat cream cheese, nine of medium-fat and 15 of high-fat whereas six reported no particular cheese habits. Procedure
Assessors received 100 g of the cheese (either low- or medium- or high-fat) and were instructed to sweeten the product ad libitum by adding sucrose and then to eat as much as they wished and to rate it by a mark on a 10-cm horizontal line, anchored at the left-hand by "ideal for me" (in French) and at the right-hand by "very far
103
INFLUENCE OF CONTEXT ON ACCEPTABILITY
TABLE 1
Effect (Ai) of label on distance from ideal for each user group at each fat level A ~ ( c m ) SEM
User group
n
Product
Low
29
Medium
12
High
30
Low fat + 0.45 Medium fat +0.90 High fat +0.04 Low fat +0.21 Medium fat +0.43 High fat + 1.35 Low fat -0.72 Medium fat -0-10 High fat + 1.51
0-46 0.54 0.44 0.72 0.61 0.58 0.47 0-38 0.55
p 0-3 0-1 0"9 0"8 0.5 0.04 0.14 0.8 0.009
from my ideal of cream cheese". It was possible to use such anchors as this product was a familiar product to all the consumers who participated in these tests.
Data analysis The difference (A~) between the distance from ideal given to the blind product and the distance from ideal given to the labelled product and the difference (Az) between the distance from ideal given at home and the distance from ideal given in the laboratory were calculated for each consumer and each type of cream cheese. Analyses of variance were performed on the data of the consumers who reported being regular users of one of the three products, respectively 71 for experiment 1 and 46 for experiment 2, in order to determine if fat levels and/or user groups should be analysed separately. The model was user group + subject (user g r o u p ) + fat + user group × fat, with the denominator for user group being subject (user group). The appropriate paired t-tests were then performed to see if A~ or A2 was significantly different from zero.
RESULTS
Experiment 1 The different user groups seemed not to react to the label in the same way at the different fat levels; user group x fat interaction F(4, 136) = 2.11; p = 0"08. Consumers of medium- and high-fat cream cheese appeared to score the high-fat cream cheese closer to ideal when it was evaluated in its commercial packaging than when without label (Table 1).
Experiment 2 The results of the analysis of variance revealed a non-significant interation between user group and fat. The different users groups of consumers reacted in the same way to the difference in location, F(4, 86) = 1-09. However, this effect of location
104
B. DA1LLANT-SPINNLERAND S. ISSANCHOU
TABLE 2 Effect (A2) of location on distance from ideal for each fat level
Product
n
A2(cm) SEM
Low fat Medium fat High fat
52 52 52
+0.41 +0-11 + 1'17
0.46 0"39 0.36
p 0.4 0.8 0.002
seemed to differ according to the fat level, F(2, 86)= 2.14, p<0-12: consumers were more severe on the high-fat cheese at home than in the laboratory (Table 2).
DISCUSSION The results of experiment 1 were consistent with those of Hellemann et al. (1990) in that the product which was preferred in blind testing became more highly regarded in labelled testing whatever the user group. The interpretation given by Cardello and Mailer (1982) could be applied here: according to these authors "the food is never as good or as bad as the mental image". High-fat users had a positive image of the high-fat product so their ratings were higher when it was presented with its label than when without, whereas they seemed to underestimate the low-fat cream cheese. On the other hand, low-fat users probably had a bad health image of the high-fat product and so they did not overestimate it. In the second experiment, only ratings of the high-fat product were affected by the location effect. An explanation could be that the location affects only the hedonic dimension and this is particularly important for the high-fat product and of little importance for the low-fat product, whatever the user group. It could be argued that assessors overestimated the products in the laboratory because during home consumption consumers refer to a broader range of products (i.e. other brands). Less favourable ratings in home or canteen environments than in the laboratory have been also reported by Hellemann et al. (1992). Another possibility is that assessors like to please the experimenter when they rate the products in the laboratory. These two studies illustrate how context can interact with the sensed fat level and the consumers' habits and underline the importance of segmenting consumers in investigations of influences on acceptability. Evidence is now needed on the processes that mediate such effects of label and location on the acceptability of food.
REFERENCE NOTES 1. Hellemann, U., Mela, D. J., Aaron, J. I. & Evans, R. E. (1992) Role of fat in meal acceptance. Presented at R. M. Pangborn Memorial Symposium on Advances in Sensory Food Science, Helsinki, Finland, 2-6 August. 2. Daillant-Spinnler, B. & Issanchou, S. (submitted for publication). Influence of fat content on acceptability of cream cheese: ideal relative response, consumption and attitudes.
INFLUENCE OF CONTEXT ON ACCEPTABILITY
105
REFERENCES Aaron, J. I., Mela, D. J. & Evans, R. E. (1994) The influences of attitudes, beliefs and label information on perceptions of reduced-fat spread. Appetite, 22, 25-37. Cardello, A. V. & Mailer, O. (1982) Relationships between food preferences and food acceptance ratings. Journal of Food Science, 47, 1553-7. Eiser, J. R., Eiser, C., Patterson, D. J. & Harding, C. M. (1984) Effects of information about specific nutrient content on ratings of "goodness" and "pleasantness" of common foods. Appetite, 5, 349-59. Hellemann, U., Tuorila, H., Lampi, A. M. & Matuszewska, I. (1990) Hedonic responses and attitudes related to fats used as spreads on bread. Food Quality and Preference, 2, 29-38. K6ster, E. P. (1981) Sensory evaluation in a "natural environment". In P. Schreier (Ed.), Flavour, 81. Pp. 93-100. Berlin: Walter de Gruyter. K6ster, E. P., Beckers, A. W. J. M. & Houben, J. H. (1987) The influence of health information on the acceptance of a snack in a canteen test. In M. Martens, G. A. Dalen, H. Russwurm (Eds.), Flavour science and technology. Pp. 391-8. Chichester: John Wiley & Sons. Light, A., Heymann, H. & Holt, D. L. (1992) Hedonic responses to dairy products: effects of fat levels, label information, and risk perception. Food Technology, 46(7), 54--7. Mailer, O., Dubose, C. N. & Cardello, A. V. (1980) Consumer opinions of hospital food and food service. Journal of The American Dietetic Association, 76, 236-42. Meiselman, H., Hirsc, E. S. & Popper, R. D. (1988) Sensory, hedonic and situational factors in food acceptance and consumption. In D. M. H. Thomson (Ed.), Food acceptability. Pp. 77-87. London: Elsevier. Rozin, P. & Tuorila, H. (1993) Simultaneous and temporal contextual influences on food acceptance. Food Quality and Preference, 4, 11-20. Solheim, R. (1992) Consumer liking for sausages affected by sensory quality and information on fat content. Appetite, 19, 285-92.
Received 19 November 1993, revision 10 March 1994
Influence of Label and Location of Testing on Acceptability of Cream Cheese Varying in Fat Content
Bi~ATRICE DAILLANT-SPINNLER and SYLVIE ISSANCHOU Laboratoire de Recherches sur les Ar6mes, INRA, Dijon
The acceptability of low-, medium- and high-fat versions of a cream cheese to consumers was measured in different conditions by rating overall distance from ideal. The influence of label (unlabelled vs. commercial package) differed according to the cream cheese fat-level habits of the assessors and to the fat level of the product. In labelled testing, high-fat cream cheese was more highly accepted by high- and medium-fat users and low-fat cream cheese seemed to be further from ideal for high-fat users. The location of testing (home vs. laboratory) affected the rating of only the high-fat product: consumers are more severe on it at home than in the laboratory.
INTRODUCTION Many authors underline the importance of context as an influence on food choice (e.g. K6ster, 1981). This paper deals with the influences on acceptability of a cream cheese by two non-food contextual influences (as defined by Rozin & Tuorila, 1993): labelled fat content and location of testing, in the home and in a laboratory. With respect to effects of information on fat content, Eiser et al. (1984) found that children (13-14 years) rated food high in fat content less pleasant than food low in fat content when information was given, whereas the contrary was observed with control subjects who received no information. K6ster et al. (1987) found that the relative amount consumed of a low-fat product (vegetable rolls) was less when health information was given than when no information was given. Solheim (1992) described for sausages an effect of fat-content information depending on the fat level and/or the sensory difference between low-fat and high-fat versions. Hellemann et al. (1990) reported that high-fat spreads were better liked in labelled testing. Light et al. (1992) reported tendencies (p<10%) for hedonic responses to full-fat and lowfat ice cream to be affected by nutrition information: ice cream with information may have been liked more than ice cream presented without information. Thus information on fat content may increase or decrease acceptability; one explanation for differences in effect could be that the way information on fat content influences acceptability depends on the attitudes of the assessors towards health. Such interaction between consumer attitudes and information on the product was considered by The authors would like to thank P. Schlich for his help in planning the experiment and analysing the data. Address correspondence to: B. Daillant-Spinnler, 1NRA, Laboratoire de Recherches sur les Ar6mes, 17 Rue Sully, BV 1540, 21034 Dijon Cedex, France. 0195-6663/951020101 +05 $08.00/0
© 1995 Academic Press Limited
102
B. DAILLANT-SPINNLERAND S. ISSANCHOU
Aaron et al. (1994): according to these authors, information leads to a shift of hedonic judgments in a direction more consistent with attitudes of the subject. Regarding effects of location of testing, Mailer et al. (1980) reported that military hospital patients eating on the wards expressed more favorable opinions about the food than those eating in the dining room. Meiselman et al. (1988) compared two studies, one in the field and another one in a dining room, with two different groups of soldiers but the same three rations per day. There was a difference of consumption between the groups which these authors attributed to the situational differences. However, although soldiers ate more in the dining room, pleasantness seemed higher in the field (a mean of 7.05 vs. 6"05 in the dining room). In a recent study (Note 1), the items of a meal were rated more pleasant in the laboratory in the first week of the study than in a canteen in weeks 2-5. As it seems that acceptability depends not only on the context but also on the interactions between the context and characteristics of the food or of the consumer, the aim of this work was to pay particular attention to the interaction of the test context with fat level and with consumer habits. The two experiments presented here are part of a larger study whose other results are to be reported separately (Note 2).
METHOD Products
Three types of the same brand of cream cheese (i.e. a dairy product close to cottage cheese but with a smooth texture) were presented to the consumers, with fat contents of 10%, 20% and 40% of dry weight (or 1.5%, 3% and 8% of total weight). Design
Experiment 1 was conducted to study the effect of label on the distance from ideal. The consumers tested the products first blind and then in their commercial package. Only fat content was indicated on the commercial packaging with a color corresponding to each fat content (pink for 10%, blue for 20% and green for 40%). It should be noted that, in France, this color code is the same whatever the brand of cream cheese. Eighty-two consumers participated to this test; 29 reported at the beginning of the study being regular users of low-fat cream cheese, 12 of mediumfat and 30 of high-fat whereas 11 reported no particular cheese habits. In experiment 2, the same products were presented in blind tests in the laboratory and at home. Fifty-two consumers participated in this test; 22 reported being regular users of low-fat cream cheese, nine of medium-fat and 15 of high-fat whereas six reported no particular cheese habits. Procedure
Assessors received 100 g of the cheese (either low- or medium- or high-fat) and were instructed to sweeten the product ad libitum by adding sucrose and then to eat as much as they wished and to rate it by a mark on a 10-cm horizontal line, anchored at the left-hand by "ideal for me" (in French) and at the right-hand by "very far
103
INFLUENCE OF CONTEXT ON ACCEPTABILITY
TABLE 1
Effect (Ai) of label on distance from ideal for each user group at each fat level A ~ ( c m ) SEM
User group
n
Product
Low
29
Medium
12
High
30
Low fat + 0.45 Medium fat +0.90 High fat +0.04 Low fat +0.21 Medium fat +0.43 High fat + 1.35 Low fat -0.72 Medium fat -0-10 High fat + 1.51
0-46 0.54 0.44 0.72 0.61 0.58 0.47 0-38 0.55
p 0-3 0-1 0"9 0"8 0.5 0.04 0.14 0.8 0.009
from my ideal of cream cheese". It was possible to use such anchors as this product was a familiar product to all the consumers who participated in these tests.
Data analysis The difference (A~) between the distance from ideal given to the blind product and the distance from ideal given to the labelled product and the difference (Az) between the distance from ideal given at home and the distance from ideal given in the laboratory were calculated for each consumer and each type of cream cheese. Analyses of variance were performed on the data of the consumers who reported being regular users of one of the three products, respectively 71 for experiment 1 and 46 for experiment 2, in order to determine if fat levels and/or user groups should be analysed separately. The model was user group + subject (user g r o u p ) + fat + user group × fat, with the denominator for user group being subject (user group). The appropriate paired t-tests were then performed to see if A~ or A2 was significantly different from zero.
RESULTS
Experiment 1 The different user groups seemed not to react to the label in the same way at the different fat levels; user group x fat interaction F(4, 136) = 2.11; p = 0"08. Consumers of medium- and high-fat cream cheese appeared to score the high-fat cream cheese closer to ideal when it was evaluated in its commercial packaging than when without label (Table 1).
Experiment 2 The results of the analysis of variance revealed a non-significant interation between user group and fat. The different users groups of consumers reacted in the same way to the difference in location, F(4, 86) = 1-09. However, this effect of location
104
B. DA1LLANT-SPINNLERAND S. ISSANCHOU
TABLE 2 Effect (A2) of location on distance from ideal for each fat level
Product
n
A2(cm) SEM
Low fat Medium fat High fat
52 52 52
+0.41 +0-11 + 1'17
0.46 0"39 0.36
p 0.4 0.8 0.002
seemed to differ according to the fat level, F(2, 86)= 2.14, p<0-12: consumers were more severe on the high-fat cheese at home than in the laboratory (Table 2).
DISCUSSION The results of experiment 1 were consistent with those of Hellemann et al. (1990) in that the product which was preferred in blind testing became more highly regarded in labelled testing whatever the user group. The interpretation given by Cardello and Mailer (1982) could be applied here: according to these authors "the food is never as good or as bad as the mental image". High-fat users had a positive image of the high-fat product so their ratings were higher when it was presented with its label than when without, whereas they seemed to underestimate the low-fat cream cheese. On the other hand, low-fat users probably had a bad health image of the high-fat product and so they did not overestimate it. In the second experiment, only ratings of the high-fat product were affected by the location effect. An explanation could be that the location affects only the hedonic dimension and this is particularly important for the high-fat product and of little importance for the low-fat product, whatever the user group. It could be argued that assessors overestimated the products in the laboratory because during home consumption consumers refer to a broader range of products (i.e. other brands). Less favourable ratings in home or canteen environments than in the laboratory have been also reported by Hellemann et al. (1992). Another possibility is that assessors like to please the experimenter when they rate the products in the laboratory. These two studies illustrate how context can interact with the sensed fat level and the consumers' habits and underline the importance of segmenting consumers in investigations of influences on acceptability. Evidence is now needed on the processes that mediate such effects of label and location on the acceptability of food.
REFERENCE NOTES 1. Hellemann, U., Mela, D. J., Aaron, J. I. & Evans, R. E. (1992) Role of fat in meal acceptance. Presented at R. M. Pangborn Memorial Symposium on Advances in Sensory Food Science, Helsinki, Finland, 2-6 August. 2. Daillant-Spinnler, B. & Issanchou, S. (submitted for publication). Influence of fat content on acceptability of cream cheese: ideal relative response, consumption and attitudes.
INFLUENCE OF CONTEXT ON ACCEPTABILITY
105
REFERENCES Aaron, J. I., Mela, D. J. & Evans, R. E. (1994) The influences of attitudes, beliefs and label information on perceptions of reduced-fat spread. Appetite, 22, 25-37. Cardello, A. V. & Mailer, O. (1982) Relationships between food preferences and food acceptance ratings. Journal of Food Science, 47, 1553-7. Eiser, J. R., Eiser, C., Patterson, D. J. & Harding, C. M. (1984) Effects of information about specific nutrient content on ratings of "goodness" and "pleasantness" of common foods. Appetite, 5, 349-59. Hellemann, U., Tuorila, H., Lampi, A. M. & Matuszewska, I. (1990) Hedonic responses and attitudes related to fats used as spreads on bread. Food Quality and Preference, 2, 29-38. K6ster, E. P. (1981) Sensory evaluation in a "natural environment". In P. Schreier (Ed.), Flavour, 81. Pp. 93-100. Berlin: Walter de Gruyter. K6ster, E. P., Beckers, A. W. J. M. & Houben, J. H. (1987) The influence of health information on the acceptance of a snack in a canteen test. In M. Martens, G. A. Dalen, H. Russwurm (Eds.), Flavour science and technology. Pp. 391-8. Chichester: John Wiley & Sons. Light, A., Heymann, H. & Holt, D. L. (1992) Hedonic responses to dairy products: effects of fat levels, label information, and risk perception. Food Technology, 46(7), 54--7. Mailer, O., Dubose, C. N. & Cardello, A. V. (1980) Consumer opinions of hospital food and food service. Journal of The American Dietetic Association, 76, 236-42. Meiselman, H., Hirsc, E. S. & Popper, R. D. (1988) Sensory, hedonic and situational factors in food acceptance and consumption. In D. M. H. Thomson (Ed.), Food acceptability. Pp. 77-87. London: Elsevier. Rozin, P. & Tuorila, H. (1993) Simultaneous and temporal contextual influences on food acceptance. Food Quality and Preference, 4, 11-20. Solheim, R. (1992) Consumer liking for sausages affected by sensory quality and information on fat content. Appetite, 19, 285-92.
Received 19 November 1993, revision 10 March 1994