the use and the attributions of some sweet foods

Appetite, 1984, 5, 199-207

The Use and the Attributions of Some Sweet Foods

RITVA PRATTALA and MIRJA KEINONENDepartment of Nutrition and Department of Psychology,Division of Applied Psychology, University of Helsinki

The attributes people give to certain sweet foods and the connections between theseattributes and the use of food were assessed in 515 men and women from differentage, educational and social groups. The foods specifically named were: ice cream,sweetened yogurt, and soft drinks. The subjects were classified according to theirreported consumption frequency of the sweet foods. The attributes of foods wereassessed by the semantic differential technique. On the basis of factor analysis of theattributions, four meaning dimensions were identified for each food. The dimensions for ice cream and sweetened yogurt were labelled 'personal use', 'healthfulness', 'nutritiveness', and 'psychosocial meaning'. The mean dimensions of softdrinks were labelled 'emotional meaning', 'social meaning', 'healthfulness', and'personal use'. These meanings of the foods correlated with their reported frequencyof use: those who associated positive attributes to a particular food also used thatfood more often than the others.

During the present century, sugar consumption ha~ increased markedly in Finlandjust as in other industrialized countries. In the 1930s the average per capitaconsumption was about 25 kg per year, whilst in 1972 it was 47 kg (Sokerityoryhmanmuistio, 1976). After that the consumption has decreased about 10 kg in 10 years(Heikkila, 1981). According to some long interval predictions, sugar consumption inFinland will not decrease in the future, but will keep at the level of 35 kg per person peryear (Rouhiainen, 1979). The attitudes of Finnish health educators and nutritionalexperts to increasing sugar consumption have been predominantly negative. Finnishnutrition recommendations and other statements on nutrition have repeatedlybrought up the principle of replacing sugar and fat in the diet by other food itemshaving a higher nutritive value. In order to prevent dental caries there are alsorecommendations to reduce the frequency of using sugar (Keys, 1968; Report of theFinnish Nutrition Committee, 1981; Sokerityoryhman muistio, 1976). These sameprinciples have been evident in recent years in practically all Western nutrition policyprogrammes and national food pattern recommendations given in industrializedcountries (see e.g., Recommended Dietary Allowances, 1980). The earlier Finnishstudies have shown differences between social classes in total consumption of sugar.The highest consumption of sugar seems to take place in the lowest social groups,particularly with farmers (Seppanen, 1980). In spite of the fact that sugar and foodscontaining sugar have been particularly emphasized in health education, and effortshave been made to control their use in other ways too, these everyday foods still form an

Reprint requests and other inquiries to: Ritva PnltHilii, Department of Nutrition, University of Helsinki,SF-00710, Helsinki 71, Finland.

0195-6663/84/030199 +09 $03·00/0 © 1984 Academic Press Inc. (London) Limited

200.. ....

R. PRATTALA AND M. KEINONEN

important part of the diet and even their partial replacement by other foods might bedifficult. People have highly conflicting opinions about sweet foods (Sellerberg, 1976).Sweet foods may also have many social and psychological meanings. This study wascarried out to ascertain what kind of associations people have with their sweet foods,and what importance is attached to healthfulness in relation to other attributes. Theaim was to explore the attributes (connotative meanings) of three sweet foods: icecream, sweetened yogurt and soft drinks. Information on the frequency of use of icecream, yogurt, and soft drinks was utilized to clarify the possible connections betweenthe use and the attributes offood. For example, did the daily user of soft drinks evaluatethem differently from a person who did not use soft drinks so frequently, and was itpossible to discriminate between those who avoided sweet foods and others accordingto the attributes they gave to these foods?

The foods selected for the study of attributions-sweetened yogurt, ice cream, andsoft drinks-had potential for variation in the attitudes the respondents held towardsthem. Also, the average consumption of sweetened yogurt, ice cream, and soft drinkshas increased greatly in Finland over the last few decades. These sweetened foods donot belong to the traditional Finnish food pattern and did not become general untilWorId War II and after.

SUBJECTS

A sample of 213 Finnish men and 302 women was selected from different age,educational and social groups, the sample providing a wide range of frequencies of theuse of sugar and single sugar-containing foods. The percentages of the user types areshown in Table 1, but the other characteristics of the research material has beendescribed in earlier research reports (PrattaUi, Rasanen & Ahlstrom, 1980; Prattala,1983).

PROCEDURE

The data were collected by a questionnaire given in 1978. Details of the researchprocedure have been described earlier in Prattala et al. (1980).

TABLE 1

Percentages o.f user types of ice cream, sweetened yogurt and softdrink (n=515)

Ice SoftUser types cream Yogurt drink

Occasional user 25 30 46(once a month or more seldom)Moderate user 70 52 45(twice a month to twice a week)Daily user 5 18 13(once a day or more often)Total 100 100 100

USE AND ATTRIBUTIONS OF SWEET FOODS 201

Food Use Frequencies

Sugar consumption was not measured quantitatively. The object was only todetermine the relative position ofeach individual in the total population in this respect.The validity of variables as indicators of the actual consumption of sugar was notexamined in this study. The frequency of the use of sweetened yogurt, ice cream, andsoft drinks was assessed on six response categories (Table 1), enabling the respondentsto be classified into three types.

Food Attribute Ratings

The attributes of ice cream, sweetened yogurt, and soft drinks were measured by thesemantic differential technique (Osgood, Suci & Tannenbaum, 1957). This method hasbeen used to reveal the attributes of foods (Fewster, Bostian & Powers, 1973; Jellinek,1973; Worsley & Leitch, 1981). Kelly's repertory grid technique has recently been usedto study food belief systems of individuals (Bell, Stewart, Radford & Carney, 1981). Theitems used in this study consisted of pairs of polar adjectives or phrases (cheapexpensive, healthy-unhealthy). The items were scored on a five point scale. In thepreliminary test, items formulated by Fewster et ale (1973) were used. The final testconsisted of 22 items, 17 ofwhich were selected for further analyses (Figure 1). Deleted

I often think about this food

I frequently use this food

I frequently ta Ik about this food

Good taste

Appetizing

Not fatteni ng

Not bad for teeth

Healthy

Light

Safe

High nutritive value

High satiety value

Cheap

Important

Li ked by almost everybody

Liked by many of my friends

Gives pleasure

I rarely think about this food

I rarely use this food

I rarely talk about this food

Bad taste

Unappetizing

Fattening

Bad for teeth

Unhealthy

Heavy

Dangerous

Low nutritive value

Low satiety value

Expensive

Unimportant

Disliked by almost everybody

Disliked by many of my friends

Does not give pleasure

2 3 4 5

FIGURE 1. The average ratings (1-5) of ice cream, yogurt and soft drink measured by thesemantic differential (n = 515). The symbols denote; ice cream (.), yogurt (.) and soft drink (.).

202 R. PRATTALA AND M. KEINONEN

items had no discriminating power, had very low correlations with the rest of the items,and could not be classified as positive or negative in relation to the foods under study.

Statistical Analyses

The underlying meaning dimensions of sweet foods were isolated and identified bymeans of factor analysis. The grouping of items around a smaller number of factorsassisted in making the information abou~ semantic differential results more concise.The four factor solution of Varimax-rotated factors was taken as the basis ofinterpretation. This was done because the Eigen values were below 1·0 after the thirdfactor, and the communalities did not increase significantly after four factors. The fourfactors could also be meaningfully labelled. Factor scores were calculated to expressindividual differences in the strength of meanings. The differences in the strength ofmeanings in different user-groups were evaluated and interpreted by means ofStudent's two-tailed t-test and discriminant analysis.

RESULTS

Meaning Dimensions of Ice Cream, Sweetened Yogurt, and Soft Drinks

According to the results of the factor analysis, the meaning structures of sweetenedyogurt and ice cream resembled one another, but the meaning structure of soft drinksdeviated from these. The meaning dimensions of ice cream and sweetened yogurt were

TABLE 2Meaning dimensions of ice cream by factor analysisa

Dimension Item Loading

Personal useEigen value 3·09

HealthfulnessEigen value 1·99

NutritiousnessEigen value 1·61

Psychosocial meaningEigen value 1·36

I often think about this foodI frequently use this foodI often talk about this foodImportant

Not fatteningNot bad for teethHealthySafeLight

HealthyHigh nutritive valueHigh satiety valueImportantLiked by many of my friendsGives pleasureGood tasteDelicious

1·010·950·900·41

0·620·720·590·450·53

0·530·740·620·420·420·530·450·52

Sum of communalities 8·04Percentage of total variance 85·6

Notes: a Varimax rotation offour factors (factor loadings above 0·40).b Naming only the positive end of each scale.

USE AND ATTRIBUTIONS OF SWEET FOODS

TABLE 3Meaning dimensions of sweetened yogurt by factor analysisa

203

Dimension



NutritiousnessEigen value 1·58

Psychosocial meaningEigen value 2·58

Itemb

I often think about this foodI frequently use this foodI often talk about this foodGood tasteDelicious

Not fatteningNot bad for teethHealthySafeLight

High nutritive valueHigh satiety valueImportantCheap

Liked by many of my friendsGives pleasureGood tasteDeliciousLiked by almost everybody

Loading

0·920·860·750·510·46

0·700·770·710·410·42

0·470·520·620·42

0·600·580·760·780·61


Notes: a Varimax rotation of four factors (factor loadings above 0'40).b Naming only the positive end of each scale.

labelled 'personal use', 'healthfulness', 'nutritiveness', and 'psychosocial meaning'(Tables 2 and 3). The meaning dimensions of soft drinks were labelled 'emotionalmeaning', 'social meaning', 'personal use', and 'healthfulness' (Table 4).

The 'personal use factor' for all sweet foods, consisted of items which described thefrequency of use, thinking of and talking about the food. Items describing the food'simportance, taste and nutritive value had casual and relatively low loadings on thisfactor. The 'healthfulness factor' was quite similar for all foods. Independently of food,the items 'fattening-not fattening', 'bad for teeth-not bad for teeth', 'healthyunhealthy', 'safe-dangerous' had high loadings on this factor. 'Light-heavy' associatedwith the healthfulness of ice cream and yogurt but 'important-unimportant' with thatof soft drinks. The subjects thus agreed which attributes were typical of healthful foodsor, on the contrary, what was characteristic of unhealthful foods (Tables 2, 3 and 4).

It was possible to label a specific 'nutritiousness factor' for ice cream and sweetenedyogurt, but not for soft drinks. This may be due to the fact that there was not enoughvariation in the ratings of soft drink's nutritiousness. Soft drink was quite unanimouslyconsidered less nutritive than ice cream and yogurt. Items which had high loadings onthe factor ofnutritiousness described the meaning of sweetened yogurt and ice cream asfood and their nutritive value and energy content. The item 'healthy-unhealthy' washighly loaded on the nutritiousness factor of ice cream, the item 'cheap-expensive' onthe nutritiousness factor of sweetened yogurt (Tables 2 and 3).

Ice cream and sweetened yogurt had a quite similar 'psychosocial meaning factor',which consisted of ratings of taste, deliciousness, and opinions of friends about them,

204.. ....


TABLE 4Meaning dimensions of soft drink by factor analysisa

Dimension Itemb Loading

0·910·850·820·490·410·41

0·570·650·690·540·55

0·620·870·90

0·640·70

I often think about this foodI frequently use this foodI often talk about this foodImportantHigh nutritive valueHigh satiety value

Not fatteningNot bad for teethHealthySafeImportant

Gives pleasureGood tasteDelicious

Liked by many of my friendsLiked by almost everybody



Emotional meaningEigen value 2·38

Social meaningEigen value 1·36


Notes: a Varimax rotation of four factors (factor loadings above 0·4).b Naming only the positive end of each scale.

and pleasure brought by ice cream or sweetened yogurt. If the subject expressed that hehimself liked ice cream or sweetened yogurt, he thought that other people, at least hisfriends, liked these foods too. Or on the contrary, if he did not like ice cream or yogurt,he did not think that other people liked them either (Tables 2 and 3). In the case of softdrinks the 'emotional and social meaning factors' were independent of each other. Itwas thus possible that the subject rated the taste of soft drinks as positive, although hemight suppose that his friends had a different opinion.

Meaning Dimensions in Relation to the Frequency of Use

On the basis of the frequency ofuse, the foods of this study had different importancein the diets of the subjects. There were more moderate users of ice cream than of theother foods and fewer daily users of ice cream. There were more occasional users of softdrinks than of the other foods (Table 1). The meaning dimensions of ice cream,sweetened yogurt, and soft drinks were in accordance with the frequency of use, so thatsubjects who rated these foods as positive also used them more frequently. Thedirection of the differences was similar in the meaning dimensions of all foods, thoughall differences were not statistically significant. The differences were least significantbetween moderate and daily users (Table 5). Student's t-test and discriminant analysisgave similar results concerning the differences between the user types and the meaningdimensions. It was possible on the basis of meaning dimensions to identify anddiscriminate the daily users from the occasional users but not from the middle group,the moderate users.

USE AND ATTRIBUTIONS OF SWEET FOODS

TABLE 5The strength of meanings of sweet foods in different user groups

205

Significance of differencesFactor scoresa between groups

Occasional Moderate Daily Occasional Occasional ModerateMeaning dimensions users users users vs. daily vs. moderate vs. daily

Ice cream (n = 130) (n = 361) (n=25)Psychosocial meaning 532 490 472 x x x x x NSHealthfulness 513 496 498 NS NS NSPersonal use 579 480 381 x x x x x x x x xNutritiousness 519 495 481 NS NS NS

Yogurt (n= 153) (n = 270) (n = 91)Psychosocial meaning 533 486 486 x x x x x x NSHealthfulness 532 495 464 x x x x x x x xPersonal use 560 488 436 x x x x x x x x xNutritiousness 505 502 484 NS NS NS

Soft drink (n = 218) (n= 232) (n= 64)Emotional meaning 533 481 453 x x x x x x NSHealthfulness 510 505 445 x x NS x xPersonal use 550 481 396 x x x x x x x x xSocial meaning 528 485 461 x x x x x x NS

Notes: a The greater the factor score the more negative the meaning (e.g., 'unhealthy', 'I seldom eat/drink it', 'I do notlike it', 'bad taste', 'low nutritive value').

bTested by Student's t-test, symbols: x x x =p<O'OOl, x x =p<O'Ol, NS=p>O·Ol.

The relationships between the frequency of use and the meanings varied accordingto food and to meaning dimension. The largest differences were in the ratingsconcerning 'personal use'. The subjects thus seemed to be able to evaluate whether theyused these foods more, less, or as much as the others (Table 5). The 'psychosocialmeaning' of ice cream, sweetened yogurt, and the 'emotional' and 'social meanings' ofsoft drinks were more important and more positive to the daily users than to theoccasional users, whereas the ratings of the 'psychosocial meaning' by the moderateusers and the daily users did not differ from each other as much (Table 5).

The differences between the user types were least significant for the dimensions of'healthfulness' and 'nutritiousness'. The ratings of healthfulness and nutritiousness byfrequent users and infrequent users did not differ clearly from each other (Table 5).Nutritiousness did not seem to be a central attribute determining the use of ice creamand yogurt, because the nutritiousness factor did not discriminate the frequent users ofice cream or yogurt from infrequent users. This result was obtained in spite of the factthat at least yogurt was on the average perceived as nutritive (Figure 1). Yogurt differedfrom other foods studied here in that healthfulness accounted for its use frequency morethan for other foods. This result may have connections with the ways yogurt has beenmarketed in Finland. Sweetened yogurt is exhibited in food stores together withunsweetened milk products, and advertized as a healthy snack, for example, forchildren. Because the nutritiousness factor could not be identified for soft drinks it wasimpossible to draw conclusions about connections between frequency of use andattributions related to the nutritive value of soft drinks. Soft drinks are not usuallyregarded as food, but are used when thirsty or together with foods.

206.. ....


DISCUSSION

The three sweet foods explored in this study did not have a similar meaningstructure. The attributes of ice cream and sweetened yogurt were alike and differedfrom those of soft drinks. The foods included in this study were examples of sweetsnacks, not of ordinary meals. Therefore, increasing the variety of foods might increasethe variety ofmeaning structures too. For example, the nutritiousness factor could havemore importance in ratings of meat dishes and other components of main pleals.Correspondingly, the attributes of drinks may differ from the attributes of foods.

The extremes of user types (daily users and occasional users) could be discriminatedby the meaning structure. For the daily users of the foods, the positive emotional, social,and psychosocial meanings were more important than they were for the other usertypes. These results are not sufficient to make causal inferences about the relationshipbetween use behaviour and food attributes; the aim was not to prove whether the use ofice cream determined the meaning of it or vice versa. It is obvious that psychosocialmeanings, like the consumption patterns offoods, develop in interaction with the socialenvironment. Some studies have shown correlations between the frequency of use andthe evaluation of foods in the context of personal constructs for foods (Bell et al., 1981)and between the frequency of use and the taste, preference or liking for a food (Lau,Hanada, Kaminskyj & Krondl, 1979; Schutz & Wahl, 1981). In spite of the fact that it iseasy to see that emotional and social factors are important in food choice, it is difficultto determine clearly how strong their effect is on specific food selection situations.Thoughts, cognitions, attitudes and affects associated with the health effects and thenutritive value offoods certainly have an influence on food choice. On the other hand, itmust always be kept in mind that healthfulness is only one part of people's food beliefsystem. The importance ofhealth and the nutritive value varies greatly according to thefood, the eating situation, the person or individual, and the culture etc. (see e.g., Bell etal., 1981; Jellinek, 1973; Lau et al., 1979; Worsley & Leitch, 1981).

The connections between the nutritiousness and the healthfulness factors and thefrequency of use were not as evident as the connections between the psychosocial andthe emotional meaning factors and the frequency of use. Some confusion may havecontributed to this, if concepts of healthfulness and nutritiousness overlapped in thesubjects' minds. However, sweetened yogurt differed from the rest of the foods in thatthe frequent users of yogurt regarded it as healthful. These results illustrate thedifficulties faced by nutrition educators who arim to change people's knowledge andbeliefs about the health aspects of foods. Some sweet foods such as sweetened yogurtmay be eaten because they are perceived by many people as healthful. On the otherhand, some sweet foods such as soft drinks may be taken in spite of the fact that they aregenerally believed not to be good for the health.

REFERENCES

Bell, C., Stewart, A., Radford, A., & Cairney, P. A method for describing food beliefs which maypredict personal food choice. Journal of Nutrition Education, 1981, 13, 22-36.

Fewster, W., Bostain, L., & Powers, R. Measuring the connotative meanings of foods. HomeEconomics Research Journal, 1973, 2, 44-53.

Heikkila, T. Ravintotaseet. Maatalouden taloudellisen tutkimuslaitoksen tiedonantoja No 82.-Helsinki, -1981. (Food -Balance- Sheet's. -The-Agricultural- Economics--Research-Institute,Finland. Research Reports. In Finnish:)

USE AND ATTRIBUTIONS OF SWEET FOODS 207

Jellinek,1. The meanings of flavors and textures. Food Technology, 1973, 27, 46-55.Keys, A. Official collective recommendation on diet in the Scandinavian countries. Nutrition

Reviews, 1968, 26, 259-263.Lau, D., Hanada, 0., Kaminskyj, 0., & Krondl, M. Predicting food use by measuring attitudes

and preference. Food Product Development, 1979,13, No.5, 66-72.Osgood, C., Suci, G., & Tannenbaum, P. The measurement oj meaning. Urbana: University of

Illinois Press, 1957.PratHiHi, R. Konsumtionen av sota livsmedel i fyra befolkningsgrupper. Niiringsjorskning, 1983,

27, 26-29 (In Swedish).PrattaHi, R., Rasanen, L., & Ahlstrom, A. Consumer behaviour in relation to sweetness of food.

In P. Koivistoinen & L. Hyvonen (Eds.), Carbohydrate sweeteners. Pp. 111-125. London:Academic Press, 1980.

Recommended Dietary Allowances. Committee on Dietary Allowances. Food and NutritionBoard. National Academy of Sciences. Washington D.C., 1980.

Report of the Finnish Nutrition Committee, Ministry of Social Affairs and Health, Finland.Helsinki, 1981 (English summary).

Rouhiainen, J. Changes in demand for food items in f~l1land 1950-1977 with consumptionforecasts for 1980, 1985 and 1990. Publications dj the Agricultural Economics ResearchInstitute, No. 40. Helsinki, 1979.

Sellerberg, A-M. Om olika synsatt pa maten. Niiringsjorskning, 1976, 20, 250-262 (In Swedish).Seppanen, R. Influence ofsocioeconomic factors on dietary intake in a Finnish population study.

Nutrition in Europe. Proceedings ojthe Third European Nutrition Conference, Pp. 117-118.Stockholm, 1980.

Schutz, H., & Wahl, O. Consumer perception of the relative importance of appearance, flavorand texture to food acceptance. How man chooses what he eats. Pp. 97-116. Zurich: ForsterPublishing Ltd., 1981.

Sokerityoryhman muistio. Tyoryhmamuistio No.4 Sosiaali-ja terveysministerio. Helsinki, 1976(Report of the Finnish Sugar Committee. No.4. Ministry of Social Affairs and Health. InFinnish).

Worsley, A., & Leitch, D. Student's perceptions of favourite and disliked foods. Journal ojHuman Nutrition, 1981,35,173-187.

Received 9 September, 1983; revision 23 May, 1984

the use and the attributions of some sweet foods

Documents