Đang tải... (xem toàn văn)
food neophobia among the finns and related responses to familiar and unfamiliar foods
Food neophobia among the Finns and related responses to familiar and unfamiliar foods Hely Tuorila a, *, Liisa La È hteenma È ki a,1 , Leena Pohjalainen b , Leila Lotti b a Department of Food Technology, PO Box 27, FIN-00014 University of Helsinki, Finland b MDC Food and Farm Facts, PO Box 505, FIN-02101 Espoo, Finland Received 31 March 2000; received in revised form 6 July 2000; accepted 10 July 2000 Abstract A representative sample of the Finns (n=1083) rated the familiarity of 20 foods designated to be ``familiar'' or ``unfamiliar'' and willingness to try them. Subjects also ®lled in a 10-item questionnaire measuring their individual food neophobia. Food neophobia scores decreased with increasing education and with the degree of urbanization. Men were more neophobic than women, and the elderly (66±80 years) were more neophobic than the other age groups. Subjects with high food neophobia were less likely to have tasted or eaten the rated foods than were those with low food neophobia. Food neophobia signi®cantly predicted the willingness to try ``unfamiliar'', and also some ``familiar'' foods. In factor analysis, the items of the food neophobia scale loaded on two factors but the variance was mainly explained by the ®rst factor related to the interest in new foods. The second factor may re¯ect a general concern about trying unknown foods. Overall, the translated food neophobia scale appeared to be a valid instrument for the characterization of consumer responses to unfamiliar foods. # 2001 Elsevier Science Ltd. All rights reserved. Keywords: Food neophobia; Familiarity; Willingness to try foods 1. Introduction Food neophobia, the tendency to avoid new foods, has been regarded as a biologically meaningful mechanism that protects individuals from consuming potentially toxic foods (see Rozin & Vollmecke, 1986). Pliner and Hobden (1992) developed and validated a 10- item verbal instrument food neophobia scale (FNS) to quantify this individual trait. The items are statements rated by each individual on a 7-point scale from ``dis- agree'' to ``agree''. Canadian subjects, mainly students, served as the study population during the development and validation. The FNS has been applied in several North American and other studies related to consumer responses to unfamiliar foods in its original English form (e.g. Frank & Hursti, 1999; Frank, Reilley, Schroth, Werk & Wehner, 1997; Meiselman, Mas- troianni, Buller & Edwards, 1999; Raudenbush & Frank, 1999; Raudenbush, Schroth, Reilley & Frank, 1998; Tuorila, Meiselman, Bell, Cardello & Johnson, 1994; Tuorila, Meiselman, Cardello & Lesher, 1998); or translated in Swedish (Hursti & Sjo È den, 1997; Koivisto & Sjo È den, 1996) or in Finnish (Arvola, La È hteenma È ki & Tuorila, 1999; Pliner, La È hteenma È ki & Tuorila, 1998; Tuorila, Andersson, Martikainen & Salovaara, 1998). There has been informal concern about the meaning and interpretation of individual FNS statements in dif- ferent populations and cultures (e.g. Koivisto & Sjo È den, 1996), but the structure of the scale has not been criti- cally examined in these studies. Several studies reporting FNS values have involved actual evaluation or tasting of unfamiliar foods (e.g., Arvola et al., 1999; Tuorila et al., 1994; Tuorila, Andersson et al., 1998; Tuorila, Meiselman et al., 1998). In those studies, we have been unable to locate subjects representing the extreme end of the FNS, i.e., the most neophobic subjects. Likewise, large Swedish surveys (Koivisto & Sjo È den, 1996; Hursti & Sjo È den, 1997) suggest relatively low FNS values among adult Swedes compared to values presented by Pliner and Hobden (1992). The ®rst aim of the present study was to describe the Finnish population in terms of food neophobia, by pay- ing attention to responses from dierent demographic subgroups based on gender, age, education and living 0950-3293/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved. PII: S0950-3293(00)00025-2 Food Quality and Preference 12 (2001) 29±37 www.elsevier.com/locate/foodqual * Corresponding author. Tel.: +358-9-191-58216; fax: +358-9- 191-58212. E-mail address: hely.tuorila@helsinki.® (H. Tuorila). 1 Current address: VTT Biotechnology, PO Box 1500, FIN-02044 VTT, Finland. area. Second, to provide evidence on the factor struc- ture of the scale, responses to items were examined using factor analysis. Third, further evidence on the validity of the scale was provided by having subjects rate their willingness to try unfamiliar and familiar foods. Since the food stimuli that were rated for willingness to try were names of real foods, the eects of food neo- phobia were analyzed based on whether the subjects had tasted or eaten a product before (thus, the product should be familiar to a person) or not (in this case, the product should be genuinely unfamiliar to a person). 2. Materials and methods 2.1. Data collection The data were collected by a nationwide marketing research agency (MDC Food & Farm Facts) from respondents who regularly participate in surveys in their homes by submitting their data using a PC and a modem provided by the agency. The respondents are representative of the Finnish population. The survey was conducted during two consecutive weekends in the spring of 1996. During the ®rst weekend, subjects responded to questions on food stimuli, and during the second they ®lled in the food neophobia scale. In addi- tion to information collected for the present study, sub- jects answered other questions on their shopping and consumption habits and food attitudes (approx. 300 questions during the two weekends). Approximately 1250 respondents participated in each of the two runs of data collection, but only 1083 completed both parts. The data from these subjects were used in the analyses. A total of 20 foods were chosen to represent familiar and unfamiliar foods of both plant and animal origin (see Table 3 for food names). Along with familiar and unfamiliar ethnic foods, other types of unfamiliar foods were included: Benecol R margarine that had recently been launched in Finland as a cholesterol lowering ``functional'' food; Lo R reduced calorie chocolate bar (also manufactured in Finland); and FlavSav R tomato which was claimed to be genetically modi®ed. These three foods were included to examine how speci®cally the food neophobia scale re¯ects consumer responses to ethnic vs. other types of unfamiliar foods. The English translations of the food names, given in Table 3, corre- spond to the original Finnish expressions as closely as possible. The respondents rated the food stimuli on two 5-point scales. The familiarity scale consisted of ®ve categories labeled ``I do not recognize the product''=1; ``I recog- nize the product, but I have not tasted it''=2; ``I have tasted''=3; ``I occasionally eat the product''=4; and ``I regularly eat the product''=5. The willingness to try or use each product was rated on a 5-point scale with only the ends verbally anchored, 1=``not at all'' and 5=``extremely''. Ten statements belonging to the food neophobia scale (Pliner & Hobden, 1992, see Table 1) were rated on a 7- point scale from ``strongly disagree'' to ``strongly agree'' (all categories verbally anchored). The gender, year of birth, education (1=lowest level/minimal, 2=middle level, 3=high school completed), and living area (grouped as 1=city or large town, 2=countryside town, 3=countryside) of the respondents had been recorded in an earlier survey. 2.2. Data analyses The individual FNS values were computed as the sum of ratings given to ten statements, after the ®ve negative items had been reversed; the FNS scores thus ranged from 10 to 70. The items were factor analyzed (maximum like- lihood method, varimax rotation). Four-way analysis of variance was used to determine the main eects of gender, age, education, living area, and their two-way interactions on FNS value. Based on their year of birth, the respon- dents were divided into six age groups: 16±25; 26±35; 36±45; 46±55; 56±65 and 66±80 years. Based on FNS scores, subjects were divided into sub- groups representing low (10±22.4), medium (22.5±45.3), and high (45.4±70) food neophobia (food neophobia grouping, FNG). The cut points were at one standard deviation (11.4) from the mean 33.9, and the respective number of subjects in each subgroup was 188 (17.4%), 740 (68.3%), and 155 (14.3%). The familiarity of food stimuli at dierent levels of food neophobia was preliminarily analyzed by dichot- omizing the familiarity ratings into categories ``not recognized/tasted'' (ratings 1 and 2) vs. ``has tried/used the product'' (ratings 3±5). Subsequently, X 2 analysis (2Â3) was used to compare familiarity (tasted vs. not tasted) in the three FNG groups. Pearson's correlation coecients were computed for an overall view of the relationships between FNS and willingness to try or eat food stimuli. Two-way analysis of variance was used to determine the main eects of familiarity (the dichotomized form, see above) and FNG, and their interaction, on the rated willingness to try food stimuli. The ANOVA, rather than correlation or regression approach, was used in the analyses because (1) we were especially interested in the will- ingness to try/eat among subjects in each extreme ends of the FNS scale behavior, and (2) the verbally anchored familiarity scale did not ful®ll the criteria of a continuous scale, but on the other hand, provided a clear cut-o point (at tasted vs. not tasted) that in ear- lier studies has proven to be important for the percep- tion of unfamiliar foods (Arvola et al., 1999). Furthermore, analyses of covariance were conducted, in which the demographic variables were added into the 30 H. Tuorila et al. / Food Quality and Preference 12 (2001) 29±37 ANOVA model as covariates, one at a time, to examine whether the demographics contributed to the will- ingness ratings to the extent that they could make the food neophobia scores redundant. 3. Results 3.1. Food neophobia scale The scale items loaded mainly on two factors (Table 1). The ®rst is related to the interest (or disinterest) in trying new and ethnic foods. The second appears to relate to a general concern vs. carelessness with respect to trying unknown foods. However, most of the variance was explained by the ®rst factor, and Cronbach's alpha of the scale was 0.847. A subsequent analysis showed that the omission of the items 3, 8 and 9, which pri- marily loaded on the second factor, would have resulted in alpha 0.853, and this derived 7-item scale would have correlated with the original 10-item scale 0.942 (Pear- son's r). Thus, no essential improvement would have been obtained by the omission of individual items loading on the second factor. The mean FNS score of the entire group of respondents was 33.9 (S.D.=11.4) (Table 2). Women were less neo- phobic than men (main eect of gender, F [1, 1024]= 9.9, P=0.002). The higher the education of a subject, the less neophobic s/he was (main eect of education, F [2,1024]=8.7, P<0.001). The food neophobia increased with age, particularly with the highest age group (main eect of age group, F [5, 1024]=3.8, P=0.002). The food neophobia decreased with the increasing degree of urbanization (main eect of living area, F [2,1024]=7.0, P=0.001). No signi®cant interactions among these factors were observed. 3.2. Familiarity of food stimuli Most subjects had tried or eaten most foods desig- nated to be familiar, and most subjects had not tried nor did not recognize foods that were designated to be unfamiliar (Table 3). However, pasha, turkey and meat balls appeared to be intermediate in terms of familiarity, as roughly two-thirds of the subjects recognized and had at least tried them, while one-third had not. Pasha is a high-fat, high-sugar dairy product of eastern Finnish origin and is an essential part of the Greek Catholic Easter celebration. Turkey meat has been available in Finland mainly during the past decade. Meat balls are very familiar to the Finns, and the unexpectedly low recognition rate derives from our description that referred to a special type of meat balls. A clear relationship was observed between FNG and the percentage of subjects who had tasted the food sti- muli (Table 3). Seven foods in the group of familiar foods and six foods in the group of unfamiliar foods had been tried signi®cantly more often by those with low food neophobia than by those high in food neo- phobia. Table 1 Food neophobia scale, mean values of items and varimax rotated factor matrix. Items negative to food neophobia, marked with R, were recoded prior to analyses. Loadings that are higher on either factor have been marked bold Item Mean SD Factor 1 Factor 2 1R I am constantly sampling new and dierent foods. 4.0 1.7 0.532 0.080 2 I don't trust new foods. 2.8 1.5 0.550 0.384 3 If I don't know what is in a food, I won't try it. 3.9 1.9 0.203 0.571 4R I like foods from dierent countries. 3.0 1.7 0.755 0.201 5 Ethnic foods look too weird to eat. 3.1 1.7 0.621 0.426 6R At dinner parties I will try a new food. 2.9 1.6 0.575 0.145 7 I am afraid to eat things I have never had before. 3.1 1.7 0.545 0.548 8 I am very particular about the foods I will eat. 3.9 1.8 0.024 0.695 9R I will eat almost anything. 3.7 1.9 0.250 0.495 10R I like to try new ethnic restaurants. 3.4 1.9 0.734 0.190 % Variance explained 38.4 7.7 Table 2 Food neophobia scores by gender, age, education, and living area Variable Categories Mean SD Range N Gender Men 35.4 11.9 10±70 507 Women 32.5 10.7 10-67 576 Age (years) 16±25 32.3 10.5 11±64 141 26±35 32.3 11.1 10±67 256 36±45 33.5 11.4 10±70 251 46±55 34.2 11.7 10±68 214 56±65 34.9 11.2 15±69 137 66±80 40.2 10.6 18±70 84 Education a Low 38.4 10.8 14±70 305 Medium 33.6 11.3 11±70 429 High 30.4 10.5 10±67 344 Living area b City or large town 32.9 11.3 10±70 707 Countryside town 34.7 11.0 10±68 256 Countryside 37.9 11.9 11±70 110 a Missing cases, n=5. b Missing cases, n=10. H. Tuorila et al. / Food Quality and Preference 12 (2001) 29±37 31 3.3. Willingness to try food stimuli The correlations between FNS and willingness to try or eat a food ranged from 0 to À0.35 (Table 3). They were highly signi®cant for all ``unfamiliar'' foods and lower, yet in most cases signi®cant, for the ``familiar'' foods. An earlier contact with a food (had tasted a food vs. not) had a signi®cant main eect on the willingness to try all twenty foods (Figs. 1 and 2, Table 4). This held true regardless of the ``general'', cultural perception of the familiarity, thus subjects maintained their reluctance to taste if they had not done it earlier, even when a food was culturally very familiar. Food neophobia group (FNG) had a signi®cant main eect on the willingness to try nine foods: pineapple, strawberry yogurt, bread cheese, beef tomato, pasha, turkey, snails, falafel balls, and genetically modi®ed tomato (Figs. 1e,g,h,i and 2a,b,f,i,j). The ®rst four are foods that originally had been categorized as familiar, and the latter ®ve had been categorized as unfamiliar. For four foods Ð pineapple, strawberry yogurt, beef tomato, and snails Ð the eect was further quali®ed by interaction between familiarity and FNG. For Lo R chocolate bar and for tofu, the familiarity by FNG interaction was the only signi®cant eect. The trends shown in Figs. 1 and 2 suggest that an earlier experience of pineapple, strawberry yogurt, beef tomato, Lo R cho- colate bar and tofu increased the willingness to try among the most neophobic subjects, compared to those neophobics who had not tried these foods. A dierent interaction was observed in the case of snails, where the food neophobia status was a stronger predictor of unwillingness to try among those who had earlier experi- ence of snails, compared to those who had not tried snails. Thus, highly neophobic subjects were encouraged to try most foods based on their earlier experience, but in the case of snails, they were discouraged by their experience. Only eleven respondents reported that they had tasted the genetically modi®ed FlavSav R tomato and among these, none was in the highest FNG group. Therefore, only one-way analysis was conducted, resulting in a highly signi®cant negative relationship between the willingness to try and the increasing FNG (Fig. 2j). Table 3 Food stimuli in the order of familiarity and characterized by origin. For each food neophobia group, the percentage of those who have tried a food is given. Correlations between FNS and willingness to try are also reported. Food name a N (tasted) Origin % Subjects who have tried by food neophobia group FNS  willingness Low Medium High P b r Familiar foods Smoked ham 1048 Animal 97 97 96 n.s. g À0.07* Tuna ®sh 1037 Animal 99 97 88 *** À0.19*** Tupla chocolate bar 1022 Plant 96 95 92 n.s. À0.08** Ma È mmi c 1020 Plant 93 95 94 n.s. À0.03 n.s. Pineapple fruit 1015 Plant 97 94 89 ** À0.21*** Flora margarine 990 Plant 94 92 88 * À0.01 n.s. Strawberry yogurt 984 Animal 97 91 85 *** À0.07* Bread cheese d 958 Animal 93 90 76 *** À0.19*** Beef tomato 955 Plant 97 88 77 *** À0.23*** Meat balls (``eines'') 780 Animal 77 73 64 * 0.00 n.s. Unfamiliar foods Pasha e 764 Animal 82 73 45 *** À0.29*** Turkey cut 669 Animal 74 62 45 *** À0.25*** Lo chocolate bar f 457 Plant 51 43 30 ** À0.11*** Tofu (soy protein product) 157 Plant 27 13 5 *** À0.26*** Benecol margarine f 155 Plant 12 16 11 n.s. À0.10*** Snails 147 Animal 31 12 2 *** À0.35*** Yosa (oat bran product) 56 Plant 6 5 3 n.s. À0.20*** Longan fruit 31 Plant 4 3 3 n.s. À0.29*** Falafel balls 30 Plant 7 2 1 *** À0.26*** FlavSav tomato (genetically modi®ed) 11 Plant 1 1 0 n.s. À0.20*** a Food names are translated so as to correspond to the original Finnish expression in the questionnaire as closely as possible (familiarity cate- gories and footnote explanations were not available to subjects). b P values refer to the X 2 (df=2) tests in which the earlier tasting vs. not tasting were compared in the three FNG groups. c Finnish Easter pudding. d Finnish specialty cheese. e Eastern Finnish Easter pudding. f Nutritionally modifed brand product. g n.s.=Not signi®cant, *P<0.05, **P<0.01, ***P<0.001. 32 H. Tuorila et al. / Food Quality and Preference 12 (2001) 29±37 Fig. 1. Mean ratings of willingness to try foods originally categorized as familiar, in three groups of food neophobia (1=low, 2=medium, 3=high) by groups of subjects who had tried or eaten a product before vs. those who did not recognize or had not tried the product The values given in parentheses are the numbers of subjects in two subgroups. H. Tuorila et al. / Food Quality and Preference 12 (2001) 29±37 33 Fig. 2. Mean ratings of willingness to try foods originally categorized as unfamiliar, in three groups of food neophobia (1=low, 2=medium, 3=high) by groups of subjects who had tried or eaten a product before vs. those who did not recognize or had not tried the product. The values given in parentheses are the numbers of subjects in two subgroups. 34 H. Tuorila et al. / Food Quality and Preference 12 (2001) 29±37 Due to the associations between FNS and demo- graphics, the analyses of covariance were conducted to see if entering demographic variables as covariates would reduce the eect of FNG on willingness to try foods. Sig- ni®cant regressions were, in many cases, produced by the covariates but the signi®cant main eect of the FNG and the signi®cant interaction FNG by familiarity, obtained in the two-way analyses of variance, remained unchanged. Thus, the demographic variables were not powerful enough to supersede the impact of food neo- phobia. 4. Discussion 4.1. Food neophobia scale The food neophobia scores of the present, representa- tive group of the Finnish population were normally dis- tributed, with a signi®cant proportion of subjects in both extremes of the scale. Approximately similar mean and dispersion statistics have been observed by Pliner and Hobden (1992) in Canadian students; by Frank et al. (1997) in US students; and by Frank and Hursti (1999) in a representative US sample. On the other hand, Swedish parents had relatively low FNS values (Hursti & Sjo È den, 1997; Koivisto & Sjo È den, 1996). Sev- eral other studies on selected population groups have likewise resulted in mean FNS values lower than the present ones, e.g. in a group of UK students (Meiselman et al., 1999); in adult members of a US consumer panel (Tuorila et al., 1994; Tuorila, Meiselman et al., 1998); and in urban Finnish women who participated in cheese evaluations (Arvola et al., 1999). However, voluntary consumer panelists who know that their task will involve tasting may be a biased sample for a food neo- phobia study, even when convinced to participate (e.g. Arvola et al.). Even if only a questionnaire on foods is to be completed, there is a risk that those who complete the questionnaire hold dierent general attitudes to or interests in foods, compared to non-respondents (e.g. Hursti & Sjo È den; Koivisto & Sjo È den). In the present study, we presumably avoided a selection bias as the subjects responded to questions on foods as part of their commitment to participate in surveys on any consumer goods. The food neophobia scores were lower among women than among men, a trend that has also been observed among Swedish adults (Hursti & Sjo È den, 1997; Koivisto & Sjo È den, 1996) but not, for example, among Canadian (Pliner & Hobden, 1992) or UK (Meiselman et al., 1999) students, or among Finnish 15-year-old teenagers (Tuorila, Andersson et al., 1998). This dierence between genders was true in all age groups, as no sig- ni®cant gender by age interaction was observed. It may have cultural origins, as women get exposed to foods and various food-related issues more extensively than men, and exposures to unfamiliar foods are known to extinguish neophobic responses (Pliner, Pelchat & Grabski, 1993). An analogous ®nding is that women recognize food odors better than men do (Cain, 1982), probably because odors are more likely a part of their daily life than they are for men. Furthermore, food professionals have low food neophobia scores, com- pared to non-specialists, which further suggests that food experience in¯uences interest in trying new foods (Frank & Kalisewicz, 2000). Other demographic variables were also related to food neophobia. As to the eect of age, the high food neophobia scores among the elderly could not have been predicted by previous results. A small sample of elderly Finns had FNS scores that were close to the overall mean of the present study (Tuorila, Andersson et al., 1998), and McFarlane and Pliner (1997) reported a decreasing food neophobia with increasing age; their oldest age group consisted of subjects >40 years. A high education predicted low food neophobia, a ®nding that was similar to US and Swedish populations (Frank & Hursti, 1999). Education is likely to enhance the access and exposures to various stimuli, events and issues, and thereby it can perhaps extinguish neophobia. Table 4 Eect of food neophobia group (low, medium, high) and familiarity (not recognized or tasted vs. tried or eaten the product earlier) on willingness to try foods, 2-way analysis of variance, F ratios (df for familiarity=1; for FNG=2; for Familiarity  FNG =2; residual= 1077) Food Familiarity b FNG b Familiarity  FNG b Familiar foods Smoked ham 75.1*** 1.83 0.65 Tuna ®sh 53.0*** 0.80 0.77 Tupla chocolate bar 94.2*** 0.90 0.10 Ma È mmi 96.0*** 0.41 0.88 Pineapple fruit 30.1*** 16.2*** 9.06*** Flora margarine 146.2*** 1.06 0.25 Strawberry yogurt 86.2*** 3.25* 5.36** Bread cheese 98.3*** 4.52* 0.37 Beef tomato 92.2*** 6.04** 3.86* Meat balls (``eines'') 83.2*** 0.05 2.04 Unfamiliar foods Pasha 136.56*** 11.73*** 0.30 Turkey cut 134.99*** 13.98*** 1.96 Lo chocolate bar 100.95*** 1.69 3.51* Tofu (soy protein product) 74.40*** 1.81 4.09* Benecol margarine 57.90*** 1.41 0.12 Snails 15.21*** 36.47*** 4.39* Yosa (oat bran product) 14.40*** 1.78 0.48 Longan fruit 5.22* 2.47 0.29 Falafel balls 29.73*** 8.88*** 2.10 FlavSav tomato (genetically modi®ed) a 17.28*** a One-way analysis of variance. b *P<0.05, **P<0.01, ***P<0.001. H. Tuorila et al. / Food Quality and Preference 12 (2001) 29±37 35 The urban subjects were less neophobic than those living in the countryside, and the eect of living area may, again, be of a similar origin. Some items on the FNS scale, referring to contacts with ethnic foods and visits to ethnic restaurants, are probably irrelevant for a large majority of non-urban elderly Finns, as they do not ®t their meal patterns or life styles (Pra È tta È la È , 2000), thus leading to unreliable responses to these items. Based on the factor analysis, food neophobia items did not compose a coherent single dimension, but some of the variance was explained by a second factor related to carelessness vs. concern about eating or intake. For example, the item no. 8 (``I am very particular about the foods I will eat'') may refer to a concern caused by dietary restrictions rather than to food neophobia or neophilia; the other items loading highly on the second factor may be related to this type of rejection as well. Koivisto and Sjo È den (1996) cite unpublished data by Sjo È den suggesting, likewise, that the item 8 is an outlier; and Frank (personal communication) reports problems with the interpretation of this item among North American subjects. Subjects who are concerned about their dietary regimen may see and rate these items in that particular context. Overall, however, it is encouraging that the FNS scale operates in foreign cultures as well as it seems to do. 4.2. Responses to unfamiliar and familiar foods Whether foods were familiar or unfamiliar, less neo- phobic subjects had more frequently tasted and eaten them than more neophobic subjects. This suggests that high food neophobia is associated with a diet with fewer items and less variety than that of less neophobic indi- viduals. Hursti and Sjo È den (1997) similarly found that high food neophobia scores of subjects (Swedish mothers and fathers) were signi®cantly correlated with the number of foods that a subject had never eaten, and Raudenbush and Frank (1999) observed lower familiarity ratings of several familiar foods among neophobics, compared to neophilics. Furthermore, Raudenbush et al. (1998) reported smaller sning magnitudes of food-related stimuli among neophobics, compared to neophilics. These observations suggest that people scoring high in food neophobia are possibly not only, or perhaps not even primarily, those who have fear of new foods; they may be individuals who have little interest in foods and per- haps focus their energy and excitement on other types of issues and activities. A dramatic dierence was observed in the ratings of willingness to try among those who had at least tasted vs. those who had not tasted or even recognized a food. The willingness to try was greatly enhanced by earlier tasting. Arvola et al. (1999) likewise found that a tasting experience was highly critical for the intention to purchase a cheese, compared to attitudinal predictors. In experi- mental conditions among children (Birch, McPhee, Shoba, Pirok & Steinberg, 1987), tasting has reduced food neophobia and increased the willingness to eat. Signi®cant interactions among the familiarity and food neophobia group suggested that the tasting experience was particularly bene®cial for the highly neophobic subjects, as their willingness ratings were greatly enhanced by the earlier experience. For example, tofu and Lo R chocolate bar received high willingness ratings from highly neophobic subjects who had tasted them, while those neophobics who had not tasted rated their willingness very low. On the contrary, willingness ratings of snails were greatly discouraged among med- ium and high neophobics by the experience of having tasted snails. Thus, a positive tasting experience can be particularly encouraging for those whose fear for the novelty is high; while a negative tasting experience can discourage further experimenting among neophobics. The tasting experience helps subjects to realize whether the sensory properties are better or worse than expected (cf. Tuorila et al., 1994). Food neophobia had no eect on willingness to try the four most familiar foods (smoked ham, tuna ®sh, Tupla R chocolate bar and ma È mmi), but it signi®cantly decreased the willingness to try pineapple, strawberry yogurt, bread cheese, beef tomato, pasha, turkey, snails, falafel balls, and genetically modi®ed tomato. Most of these foods were either described as being of animal origin (strawberry yogurt, bread cheese, snails and turkey), or subjects knew or could guess this (pasha), or the name of the food was such that the origin could not be derived from it (``falafel balls''). Moreover, the name of beef tomato contains an association with meat. Only pineapple and genetically modi®ed tomato were free of associations with animal products. On the contrary, the willingness ratings were unrelated to food neophobia when the plant origin was explicit from the name (``tofu Ð soy protein product'', ``yosa Ð oat bran product'', ``longan fruit'', ``Lo chocolate bar'', ``Benecol margarine''). These responses support, at least to some extent, the ®nding by Pliner and Pelchat (1991) who reported more neophobic responses to products of animal origin than to those of plant origin; but they contrast with the ®nding by Tuorila, Meiselman et al. (1998) in which reindeer meat was rated more positively than e.g. arctic cloudberry jam when served to subjects who were unfa- miliar with these foods. In the latter study, respondents were from a consumer panel and none of them were highly neophobic. Ratings of unfamiliar, nutritionally modi®ed foods had only one signi®cant relationship with the food neo- phobia grouping, in that willingness to try Lo R chocolate bar was positively aected by familiarity in the highest neophobia group. The willingness to try Benecol R was unrelated to food neophobia. Ratings of these foods may re¯ect a non-neophobic response to nutritionally modi®ed products, but they can equally well re¯ect 36 H. Tuorila et al. / Food Quality and Preference 12 (2001) 29±37 general responses to foods of plant origin. Interestingly, the increasing food neophobia signi®cantly reduced the willingness to try genetically modi®ed tomato, although it was clearly of plant origin. Responses to Lo R choco- late bar and to FlavSav R tomato suggest that the FNS scores may predict responses to unfamiliar foods that are non-ethnic for their origin. Finally, the results strongly suggest that familiarity of a food is not a cultural, but an individual experience. A food that is well-known in a particular culture is unfamiliar for a person until s/he gets to taste it. At that point, his or her overall level of willingness to try it (again) is dramatically altered. Consequently, the only valid categorization of foods as ``familiar'' and ``unfamiliar'' can be done by individual subjects of a study. Acknowledgements This study has been partially conducted with ®nancial support from the Commission of European commu- nities, under the title ``The development of models for understanding and predicting consumer food choice'' (AIR-CT94-1315). The authors thank Dr. Hannu Rita for his useful comments on an earlier draft of the paper. References Arvola, A., La È hteenma È ki, L., & Tuorila, H. (1999). Predicting the purchase interest of unfamiliar and familiar cheese: eects of atti- tudes, liking and food neophobia. Appetite, 32, 113±126. Birch, L. L., McPhee, L., Shoba, B. C., Pirok, E., & Steinberg, L. (1987). What kind of exposure reduces children's food neophobia? Looking vs. tasting. Appetite, 9, 171±178. Cain, W. S. (1982). Odor identi®cation by males and females: predic- tions vs performance. Chemical Senses, 7, 129±142. Frank, R. A., & Hursti, U.-K. K. (1999). A cross-cultural comparison of responses on the food neophobia scale using US and Swedish samples. Poster presented at the Annual Meeting of the Society for the Study of Ingestive Behavior, Clearwater, Florida. Frank, R. A., & Kalisewicz, S. (2000). Food experience and will- ingness to try novel foods. Appetite, 34, 335. Frank, R. A., Reilley, S. P., Schroth, F. E., Werk, C. D., & Wehner, B. (1997). Food neophobia re¯ects novelty seeking, not sensation seeking, or anxiety. Paper presented at the Annual Meeting of the Society of Ingestive Behavior, Baltimore, Maryland. Hursti, U.-K. K., & Sjo È den, P.-O. (1997). Food and general neopho- bia and their relationship with self-reported food choice: familial resemblance in Swedish families with children of ages 7±17 years. Appetite, 29, 89±103. Koivisto, U.-K., & Sjo È den, P.-O. (1996). Food and general neophobia in Swedish families: parent±child comparisons and relationships with serving speci®c foods. Appetite, 26, 107±118. McFarlane, T., & Pliner, P. (1997). Increasing willingness to taste novel foods: eects of nutrition and taste information. Appetite, 28, 227±238. Meiselman, H. L., Mastroianni, G., Buller, M., & Edwards, J. (1999). Longitudinal measurement of three eating behavior scales during a period of change. Food Quality and Preference, 10, 1±8. Pliner, P., & Hobden, K. (1992). Development of a scale to measure the trait food neophobia. Appetite, 19, 105±120. Pliner, P., La È hteenma È ki, L., & Tuorila, H. (1998). Correlates of human food neophobia. Appetite, 30, 93. Pliner, P., & Pelchat, M. L. (1991). Neophobia in humans and the special status of foods of animal origin. Appetite, 16, 205±218. Pliner, P., Pelchat, M., & Grabski, M. (1993). Reduction of food neo- phobia in humans by exposure to novel foods. Appetite, 20, 111±123. Pra È tta È la È , R. (2000). North European meals: observations from Den- mark, Finland, Norway, and Sweden. In H. L. Meiselman, Dimen- sions of the meal. The science, culture, business, and art of eating (pp. 191±201). Maryland: Aspen. Raudenbush, B., & Frank, R. A. (1999). Assessing food neophobia: the role of stimulus familiarity. Appetite, 32, 261±271. Raudenbush, B., Schroth, F., Reilley, S., & Frank, R. A. (1998). Food neophobia, odor evaluation and exploratory sning behavior. Appetite, 31, 171±183. Rozin, P., & Vollmecke, T. (1986). Food likes and dislikes. Annual Review of Nutrition, 6, 433±456. Tuorila, H., Andersson, A Ê ., Martikainen, A., & Salovaara, H. (1998). Eect of product formula, information and consumer characteristics on the acceptance of a new snack food. Food Quality and Preference, 9, 313±320. Tuorila, H., Meiselman, H. L., Bell, R., Cardello, A. V., & Johnson, W. (1994). Role of sensory and cognitive information in the enhancement of certainty and liking for novel and familiar foods. Appetite, 23, 231±246. Tuorila, H. M., Meiselman, H. L., Cardello, A. V., & Lesher, L. L. (1998). Eect of expectations and the de®nition of product category on the acceptance of unfamiliar foods. Food Quality and Preference, 9, 421±430. H. Tuorila et al. / Food Quality and Preference 12 (2001) 29±37 37