ARTICLE IN PRESS
Appetite 50 (2008) 83–90 www.elsevier.com/locate/appet
Research report
Structure of the relationship between parents’ and children’s food preferences and avoidances: An explorative study Margherita Guidetti, Nicoletta Cavazza Universita` di Modena e Reggio Emilia, Reggio Emilia, Italy Received 9 February 2007; received in revised form 11 April 2007; accepted 1 June 2007
Abstract The aims of the present research were to uncover the underlying structure of the relationship between parents’ and children’s preferences and avoidances in the food domain, and to determine whether this structure revealed any differences as a function of children’s age. Two hundred and eighty-two parent–child dyads (children aged 10–20) completed a self-administered questionnaire on eating attitudes and practices. The results led to a descriptive model of the connection between parents’ and children’s food repertories, made of four overlap and four autonomy areas. This structure was then compared with the structure that emerged from random pairs of an adult and a child/adolescent. Our findings showed that parents’ repertory had an anchoring function in the formation of that of their children, not only in terms of imitation, but also in terms of influencing the direction of differentiation and innovation. As far as the second aim of our study is concerned, only two out of eight areas (that of parent preferences limitation, and that of child autonomous avoidances) differed in relation to the children’s age. The theoretical and practical impact of these results is discussed. r 2007 Elsevier Ltd. All rights reserved. Keywords: Food preferences; Food avoidances; Parents; Children; Adolescents
Introduction Aside from innate (or genetic) predispositions, individual food preferences and avoidances begin developing in childhood both through personal experience and social interaction. The former includes learning through simple and associative conditioning (e.g., Brunstrom, 2004; Wardle et al., 2003). The latter comprises social facilitation, modeling, and deliberate influence through interpersonal and media communication (for a review, cf. Conner & Armitage, 2002). In every process listed above, the role of family is substantial, since by the age of 3–4 years eating behaviour of children is increasingly influenced by the physical and social environment (Rolls, Engell, & Birch, 2000). Humans are genetically predisposed to reject novel edible foods (food neophobia) and to prefer the more familiar ones. However, repeated exposure and social influence can Corresponding author.
E-mail address:
[email protected] (M. Guidetti). 0195-6663/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.appet.2007.06.001
transform an initial neophobic response toward a new food into a preference (for a review cf. Birch, 1999). Research on parents’ influence on children’s eating habits has shown that this influence manifests itself through different processes (Birch & Fisher, 1998; Hertzler, 1983; Nicklas et al., 2001; Patrick & Nicklas, 2005): genetic transmission (e.g., Breen, Plomin, & Wardle, 2006); restriction of food experience, including the selective exposure (e.g., Cullen et al., 2001; Kratt, Reynolds, & Shewchuk, 2000) and the control, at least partial, on the physical, social and affective context of eating (e.g., Campbell, Crawford, & Ball, 2006; French, Story, Neumark-Sztainer, Fulkerson, & Hannan, 2001; NeumarkSztainer, Hannan, Story, Croll, & Perry, 2003; Videon & Manning, 2003); modeling (e.g., Addessi, Galloway, Visalberghi, & Birch, 2005; De Bourdeaudhuij & Van Oost, 2000); parental practices, that is pressures, restrictions and monitoring (e.g., Birch et al., 2001; Carper, Orlet, & Birch, 2000; Liem, Mars, & De Graaf, 2004), and parental styles related to their children’s nutrition (e.g., Kremers, Brug, de Vries, & Engels, 2003; Patrick, Nicklas,
ARTICLE IN PRESS 84
M. Guidetti, N. Cavazza / Appetite 50 (2008) 83–90
Hughes, & Morales, 2005; Vereecken, Keukelier, & Maes, 2004). This would thus suggest a relatively high degree of resemblance between children’s and parents’ food preferences. Many studies aimed at quantifying the extent of this similarity have found positive, though fairly weak, correlations between parents’ and children’s preferences, which might be explained by the shared cultural environment (e.g., Birch, 1980; Borah-Giddens & Falciglia, 1993; Ritchey & Olson, 1983; Rozin, 1991; Rozin, Fallon, & Mandell, 1984; Stafleu, Van Staveren, De Graaf, Burema, & Hautvast, 1995; Weidner, Archer, Healy, & Matarazzo, 1985). With reference to this, Rozin (1991) talked about ‘‘family paradox’’. Instead of using correlations, some authors crosstabulated the liked and the disliked foods of the members of each dyad. Pliner and Pelchat (1986) calculated a phistatistic for each table, whereas Skinner and colleagues (Skinner et al., 1998; Skinner, Carruth, Wendy, & Ziegler, 2002) computed a percentage of agreement (like–like matches+dislikedislike matches out of the potential correspondences) and found the highest similarity scores compared with other studies. However, these research were focused on resemblances and left out differences between parents’ and children’s food repertories (thought Skinner et al., 1998, reported also the percentage of correspondence between likes and dislikes of the members). In all these studies, preferences were collected through questionnaires (food preference questionnaire) or interviews in which participants were asked how much they liked each food listed (‘‘hedonic rating’’). Although the researchers used the term ‘‘preferences’’, in our opinion, these studies investigated the similarity between parents’ and children’s likes and dislikes. Indeed, authors generally use the term ‘‘preference’’ as a synonym of ‘‘liking’’. However, Rozin and Vollmecke (1986) suggest to distinguish between these expressions: the former presumes the availability of at least two different items, and implies the choice of one rather than the other; the latter indicates the affective reactions to a food. They stated that liking is only one of the motivations that may account for a preference. Other factors (e.g., perceived healthiness, convenience, price) can affect preferences but not liking. Rozin and Vollmecke (1986) proposed a taxonomy of preferences’ and avoidances’ underlying reasons: these are sensory-affective motivations, anticipated consequences, and ideational motivations. On the basis of different combinations of these three classes of motivations, Rozin and colleagues (Fallon & Rozin, 1983; Rozin & Fallon, 1980, 1987) have designated four types of food rejections: distaste, danger, disgust, and inappropriate. In line with this conceptualization, we will use the terms ‘‘preference’’ and ‘‘avoidance’’ as broader concepts than ‘‘like’’ and ‘‘dislike’’. Since research based on hedonic rating highlighted moderate correlations between parents’ and children’s likes and dislikes, it is possible that comparing them at a wider
level could yield different results. We believe that the connection between parents’ and children’s food repertories should be considered from a perspective that goes beyond the quantification of the similarity between likes and dislikes. Parental influence in the food domain could manifest itself not only in imitation, but also in the anchoring function that the parents’ repertory fulfills in the formation of their children’s repertory. Using their parents’ preferences and avoidances as a reference point, children could build their food repertory by reproducing that of their parents but also by contradicting it and expanding it. Thus, the structure of the relationship between parents’ and children’s repertories might include not only intradyad similarities and differences, but also autonomy zones of each participant. Although research in the past has largely focused on correlational analysis, correlation seems not completely suitable to highlight the structure of the relationship between parents’ and children’s food repertories. In our opinion, the use of cross-tabulations is a better way to stress this structure. Nevertheless, as the cross-tabs were based on hedonic rating task, they did not allow to consider the foods that are particularly preferred or avoided by only one member of the dyad but not by the other. Indeed, participants could give a judgement online, even if some items were not spontaneously comprised in their personal set of preferences and avoidances. In this way, every food would be included in one of the four possible intersections between likes and dislikes of the dyad members. Suppose that each individual hold a continuum from preferred to avoided foods, with a wider neutral zone in the middle: by asking participants to list the foods placed at the extremity of this continuum, we could obtain the overlaps between parents’ and children’s preferences and avoidances, as well as the autonomy zones, representing the intersection between one extremity of one member of the dyad and the neutral zone of the other. In other words, the autonomy zones include the foods that are salient for one member of a dyad but that are not salient for the other. Indeed, if parents’ and children’s repertories would not be linked, all the preferences and all the avoidances of one member of the dyad should fall in the neutral zone of the other. This would be the null hypothesis of complete independence between the two food repertories. In addition, the autonomy zones allow us to observe to what extent the child explores outside the parental repertory of the main preferences and avoidances. Furthermore, the research has left unexplored the question concerning to what extent the resemblance between parents’ and children’s likes develops in relation to the age of the latter. This is an important point, however, since likes and dislikes, preferences and avoidances change as children get older and are exposed longer to foods (Fischler, 1990; Logue, Logue, Uzzo, McCarty, & Smith, 1988). In addition, as children grow up, they modify their beliefs and attitudes about differing foods (Conner, 1993) and they are increasingly exposed to influences from
ARTICLE IN PRESS M. Guidetti, N. Cavazza / Appetite 50 (2008) 83–90
outside the family (Skinner et al., 1998). It is also necessary to bear in mind the special nature of the transitional stage from childhood to adulthood, characterized by the elaboration of identity. Adolescence is typically a time of gradual shift from parental to peer influence (Erikson, 1963). Food choice is deeply linked to the construction and expression of identity both at personal and social level (Bisogni, Connors, Devine, & Sobal, 2002; Vartanian, Herman, & Polivy, 2007), and particularly in this phase of life (like much other behaviour) it fulfills a function of selfexpression. While studies focusing on the resemblance between parents’ and children’s food preferences have not dealt with the question of its variation in relation to children’s age, the research concerning the resemblance between eating habits does not present convergent results. Some researchers (Hannon, Bowen, Moinpour, & McLerran, 2003; Patterson, Rupp, Sallis, Atkins, & Nader, 1988) reported greater similarities between parents’ and children’s food consumption when children are young rather than adolescent. Instead, De Bourdeaudhuij (1996) showed higher correlations between parents and adolescent children than between parents and younger children. Still others (Lau, Quadrel, & Hartman, 1990; Rossow & Rise, 1994) indicated a certain stability in parental influence throughout their children’s lives. Lastly, Feunekes, Stafleu, De Graaf, and Van Staveren (1997) found different correlations for food items consumed at home and those consumed outside the home: for the former, associations were higher and did not differ significantly with the increasing age of children, whereas for the latter, associations were lower and more unstable. Furthermore, while correlations were never significant for father–child dyads, they were for mother–child dyads until the child reached the age of about 12. In summary, to the best of our knowledge, studies to date have not explored two questions: the first concerns the fact that association between food likes scores of children and their parents have been surprisingly weak. Nevertheless, studies carried out in this domain were limited to analyzing correlations or cross-tabulations between hedonic rating scores of parents and children. Hence, the complexity of the relationship between food repertories was reduced to a simple problem of similarity between parents’ and children’s judgement on a like–dislike continuum. The second question concerns the possible variations in the relationship linking children’s and parents’ food repertory as a function of the age of the former. The present study had two aims. First, to uncover the underlying structure of the connections between children’s and parents’ food preferences and avoidances, made of overlap and authonomy areas. Given the exploratory nature of our study, we hypothesized that the relationship between parents’ and children’s food repertories would be stronger than the relationship linking random dyads of an adult and a child/adolescent. The second aim was to
85
determine whether this structure showed any differences in relation to the children’s age. Method Participants and procedure Two hundred and eighty-two Italian parent–child dyads volunteered for the study. Children were 150 females and 132 males aged 10–20 years (M ¼ 13.78, SD ¼ 2.84). They were recruited from Scout groups from Carpi (MO). Each child was coupled with the parent who usually did the food shopping and cooking: 238 mothers and 44 fathers aged 31–63 years (M ¼ 43.60, SD ¼ 4.79). The effect of parents’ and children’s sex was tested in all analyses: no significant effects were observed, and therefore will not be presented. Each member of the dyads separately completed a questionnaire on food attitudes and practices. For the purpose of this study, we analyzed the answers to two open questions. Instead of supplying participants with a list of foods to evaluate on a like–dislike continuum, we asked respondents to indicate the six foods placed at the two extremity of the preferences–avoidances continuum (‘‘please list three dishes or foods that you prefer’’; ‘‘please list three dishes or foods that you strongly dislike and would like to avoid’’). The open-ended questions are suitable to capture participants’ preferences and avoidances because they assume the availability of endless alternatives. Furthermore, this method has three other benefits. First, it allows to take into account foods that are particularly preferred or avoided by a member of the dyad but are not by the other (what we called ‘‘autonomy areas’’). Second, it might overcome the tendency shown by children to accept a narrower number of foods with respect to the adults (Pliner & Pelchat, 1986): indeed, this tendency does not influence the results if participants were asked to list their preferences and avoidances. Lastly, this strategy could strengthen our results, since giving a consistent judgement about foods listed by researcher is different from indicating spontaneously the same food category or even the same foodstuff among thousands of different food items that could be reported. Two hundred and seventy-six parent–child couples answered the questions and formed our final sample. Thus, we obtained information about preferences and avoidances areas for each member of the dyads. Through content analysis, preferred and avoided foods were grouped into 17 categories (pasta, traditional first courses, grains, soup, pizza and bread, meat, cold cuts, offal, fish, dairy products and eggs, pulses, vegetables, fruit, desserts, fast food, junk food and snacks, ethnic food). Two persons (one of the authors and another person) allocated each food to a category and disagreements between coders (76 out of 4584 item classified, that is 1.66%) were resolved throughout discussion. For example, the ‘‘vegetables’’ macro-category included tomatoes, lattuce, cabbage, and also potatoes;
ARTICLE IN PRESS M. Guidetti, N. Cavazza / Appetite 50 (2008) 83–90
86
‘‘dessert’’ category comprised homemade sweet items, icecream, and chocolate; while other confectionery and chips were placed in ‘‘fast food, junk food, and snack’’ category. Measures and analysis In order to uncover the underlying structure of the relationship between children’s and parents’ food preferences and avoidances, we followed a modified version of the cross-tabulation method used by Pliner and Pelchat (1986) and Skinner et al. (1998, 2002). We gathered preferences and avoidances separately in order to compare distinct overlap areas: not only resemblances (food items preferred or avoided by both child and parent), but also differences (food items preferred by one and avoided by the other). We represented preferences and avoidances of each parent and his/her child graphically as circles intersecting each other (Fig. 1). Then we measured the size of the four overlap areas by counting foods and food categories in common among the various zones (both preferences and avoidances) and assigning 2 points for any matching foodstuff (if the parent and the child indicated the same specific food) and 1 point for any matching food category (if they both indicated a food included in the same category) between the repertories of each dyad. Two persons were responsible for allocating these points. Hence, a correspondence concerning the same specific foodstuff weighed more than a correspondence at
food category level. This was an arbitrary choice, but it is justifiable because the same preference for ‘‘pasta’’ is less reliable than the same preference for ‘‘spaghetti alla carbonara’’, for example. Suppose that a child listed as preferences ice cream, pizza and apricot, while his mother listed pizza, chocolate cake and carrots. They also listed as avoindances cabbage, gorgonzola cheese and liver, and mortadella, liver and panettone, respectively. This dyad would get a score of 3 for the preferences reproduction area (1 point for the match ice cream/chocolate cake—same ‘‘dessert’’ category—+ 2 point for the match pizza/pizza—same specific foodstuff); a score of 2 for the avoidances reproduction area (for the match liver/liver), a score of 1 for the limitation area (for the match cabbage/carrots) and a score of 1 for the opposition area (for the match ice cream/panettone). Thus, the obtained overlap areas (two positive and two negative) could vary from 0 to 6. They were called
PR=1.56
the preferences reproduction (PR) area, resulting from the intersection between parent’s and child’s preferences (positive overlap), the avoidances reproduction (AR) area, resulting from the intersection between parent’s and child’s avoidances (positive overlap), the parental repertory limitation (L) area, resulting from the intersection between parents’ preferences and child’s avoidances (negative overlap), and the opposition (O) (against parental repertory) area, resulting from the intersection between parent’s avoidances and child’s preferences (negative overlap).
As we mentioned above, the open-ended question method allowed us to take into account also the autonomy areas, which includes foods that are particularly preferred or avoided by a member of the dyad but are not by the other: PAP=3.72
CAP=4.07
L=0.72
O=0.38
CAA=4.30
PAA=4.65
AR=0.97
Fig. 1. Structure of the relationship between parents’ and children’s food repertories: intersection between preferences and avoidances of participating pair members (means of the areas size). Range: 0–6. Legend: ‘‘positive’’ overlap areas: PR ¼ preferences reproduction area; AR ¼ avoidances reproduction area; ‘‘negative’’ overlap areas: L ¼ parental repertory limitation area; O ¼ opposition (towards parental repertory) area; autonomy areas: PAP: parents’ autonomous preferences area; PAA, parents’ autonomous avoidances area; CAP, child’s autonomous preferences area; CAA, child’s autonomous avoidances area.
parent’s autonomous preferences (PAP) area, resulting from the subtraction of preferences reproduction area and limitation area from the parent’s whole preferences zone (PAP ¼ total parent’s preferences(PR+L) ¼ 6PRL), child’s autonomous preferences (CAP) area, resulting from the subtraction of preferences reproduction area and opposition area from the child’s whole preferences zone (CAP ¼ total child’s preferences(PR+O) ¼ 6PRO), parent’s autonomous avoidances (PAA) area, resulting from the subtraction of avoidances reproduction area and opposition area from the parent’s whole avoidances zone (PAA ¼ total parent’s avoidances(AR+O) ¼ 6ARO), and child’s autonomous avoidances (CAA) area, resulting from the subtraction of avoidances reproduction area and limitation area from the child’s whole avoidances zone (CAA ¼ total child’s avoidances(AR+L) ¼ 6ARL).
ARTICLE IN PRESS M. Guidetti, N. Cavazza / Appetite 50 (2008) 83–90
In order to control for the effect of participants’ shared cultural environment, we adopted a method already used by Birch (1980) and Pliner and colleagues (Pliner, 1983; Pliner & Pelchat, 1986). The extent of the four overlap areas was also computed for random pairs of an adult and a child/adolescent of the same gender and in the same age group of his/her real child. These random pairs were generated by shifting down of one record the parents’ data, within sub-samples based on gender and children’s age range. These overlap areas were then compared with those of the parent–child dyads, to test if the former were smaller than the latter. Results A 2 (parent gender) 2 (child gender) 8 (areas) mixed ANOVA design with repeated measures on the third factor, and children’s age as covariate, was performed. Table 1 shows the mean size of identified areas, for both real and random pairs. Fig. 1 graphically represents the structure of the relationship between food preferences and avoidances concerning real parent–child dyads. The analysis showed a main effect for the repeated factor, F(4.268) ¼ 22.711, po.001, Z2 ¼ .253. Pairwise comparisons (Bonferroni adjustment for multiple comparisons) confirm that every area was significantly different from the others at the .001 level (except for the AR–L and the CAP–CAA pairs, which are not significantly different). The broader overlap area was that of preferences reproduction, while the smallest was the opposition area (PAA4CAACAP4PAP4PR4ARL4O). There were no interaction effects between areas and children’s age, parent gender, and child gender. The child autonomous preferences area was significantly larger than the opposition area, and the child autonomous avoidances area was significantly larger than the limitation area, showing children’s tendency to explore outside the parental repertory. Children seemed to explore parents’ preferences area more than parents’ avoidances area: among the positive overlaps (reproductions), the one that
87
reproduced parents’ preferences was broader than the one that reproduced his/her avoidances, and among the negative ones (limitation and opposition), the one that limited parents’ preferences was broader than the one that opposed his/her avoidances. Furthermore, the overlaps concerning the parents’ avoidances area were very narrow: the PAA area was the largest of all and it was considerably greater than the PAP area. The positive overlaps were greater than the negative ones (PR4L and AR4O), indicating that our participants were more inclined to imitate parental food repertories than to refuse them. An analysis of variance with repeated measures on the factor ‘‘areas’’ was performed for the random pairs as well. The results showed a main effect for the repeated factor, F(4.259) ¼ 31.134, po.001, Z2 ¼ .325: every area was significantly different from the others at list at the .01 level (except for the ARL and the CAACAP pairs, which are not significantly different). The same structure emerged from every comparison: CAP4O, po.001; CAA4L, po.001; PR4AR, po.001; L4O, po.01; PR4L, po.001; AR4O, po.001; PAA4PAP, po.001. However, as we anticipated, paired sample t-test showed that the overlapping areas between preferences and avoidances of the real parent–child dyads, with the exception of the opposition area, were significantly larger
Table 2 Correlation between areas size and children age (N ¼ 276)
Preferences reproduction area (PR) Avoidances reproduction area (AR) Limitation area (L) Opposition area (O) Child autonomous preferences area (CAP) Child autonomous avoidances area (CAA) Parent autonomous preferences area (PAP) Parent autonomous avoidances area (PAA) a
Pearson r
Sig. (2-tailed)
.005 .032 .133a .011 .001 .121a .092 .034
.938 .592 .027 .853 .986 .045 .126 .578
Correlation is significant at the .05 level (2-tailed).
Table 1 Means and standard deviation of overlap and autonomy areas both for parent–child and random dyads Parent-child dyads
Random dyads
Paired sample t-test
M
SD
M
SD
t
df
p
1.13 1.05 1.03 .60 1.42 1.29 1.21 1.41
1.11a .65b .55b .33c 4.34d 4.56e 5.01f 4.80e
.98 .82 .74 .58 1.14 1.09 1.08 1.10
5.037 4.068 2.394 1.012 5.965 5.100 3.865 4.642
260 260 260 260 260 260 260 260
.000 .000 .017 .312 .000 .000 .000 .000
‘‘Positive’’ overlap (resemblances) Preferences reproduction area (PR) 1.56a Avoidances reproduction area (AR) .97b ‘‘Negative’’ overlap (differences) Parental repertory limitation area (L) .72b Opposition (towards parental repertory) area (O) .38c Autonomy areas Parent’s autonomous Preferences area (PAP) 3.72d Child’s autonomous preferences area (CAP) 4.07e Parent’s autonomous avoidances area (PAA) 4.65f Child’s autonomous avoidances area (CAA) 4.30e
Note: In each column, means with different subscripts differ from each other by at least po.01. See paired sample t-test columns for the differences between real and random dyads.
ARTICLE IN PRESS 88
M. Guidetti, N. Cavazza / Appetite 50 (2008) 83–90
than that identified for the random dyads, t(260) ¼ 5.037, po.001 for the PR area, t(260) ¼ 4.068, po.001 for the AR area, and t(260) ¼ 2.394, po.05 for the limitation area. As far as the relationship between the structure illustrated and the age of the children is concerned, the cited mixed ANOVA design showed no overall effect due to the covariation of the children age. Nevertheless, we measured the single correlations between area sizes (both overlap and autonomy) and children age (see Table 2): two significant, though very weak, correlations emerged. The first involved the children’s age and the size of the limitation area, r(276) ¼ .133, po.05: as age increased, the number of foods that children refused among those particularly loved by their parents decreased. The second correlation involved the children’s age and the size of child autonomous avoidances area, r(276) ¼ .121, po.05: as age increased, the number of foods that children avoided and parents did not mention increased. Lastly, we checked the effects of some socio-demographic characteristics such as parents’ occupation, parents’ educational level, number of family members, and residence area, by including them as factors in a series of ANOVAs with children’s age as covariate. No significant effects were observed. Discussion Our findings allowed us to outline a descriptive model of the connection between parents’ and children’s food repertories, showing the overlaps between preferences and avoidances for the members of each dyad, and the autonomy areas. These results suggested that parental food repertories would have an anchoring function in the formation of those of the children, not only in terms of imitation, but also in terms of influencing the direction of differentiation and innovation. The comparison between the different areas indicated that the children were more inclined to imitate parental food repertory rather than to refuse it and that their food exploration was more directed outside the parents’ preferences area than towards the parents’ avoidances area, which remained relatively unexplored. The same structure emerged from the random pairs, showing that a shared cultural environment influences individuals’ preferences and avoidances. Nevertheless, the overlap areas regarding real dyads were greater than those of the random dyads (except for the opposition area). These findings indicate that this dynamic of imitation and exploration specifically concerned the parent–child relationship. Furthermore, our results supported the idea that similarities and differences linking parents’ and children’s food repertories are not simply due to a shared culture, but were, at least in part, specifically attributable to the shared family environment. This pattern of results is consistent with those of Pliner et al. (Pliner, 1983; Pliner & Pelchat, 1986), and go beyond
them, as the present study examined the multifaceted nature of the connection between parents’ and children’s preferences and avoidances. Both preferences and avoidances reproduction areas were significantly broader for the real dyads than for the random ones. Instead, the pattern pertinent to the negative overlap areas was differentiated: the limitation area was larger for the real parent–child dyads than for the random ones, while no significant differences were found for the opposition area. These findings suggest that the imitation of preferences and avoidances and the limitation of the parental repertory of loved foods were behaviours that children activated towards their parents specifically, while rebellion consisting in preferring what adults avoided could be observed towards adults in general. Moreover, the fact that the opposition area was the only one that did not differ significantly between real and random dyads additionally supported the idea that children did not push themselves to explore the parental avoidances area. It is interesting to note, however, that Birch (1980) obtained different results, probably due to the nature of the task used in her study (ranking of a limited number and variety of foods). Our findings differed in part from those reported by Skinner et al. (1998). In their research, for both mothers and fathers, the like–like agreement is the largest one (substantially different from the others), as it is the PR area in our study. Nevertheless, the other three overlaps were exactly in the opposite order compared with our findings (child likesparent dislikes4child dislikesparent likes4both dislikes VS AR4L4O). This difference is probably due to the research method adopted. We could expect that, using a traditional hedonic rating questionnaire, overlap areas size is dependent on the food items selected by researcher, while the open-ended question method we used may provide more stable results. Regarding the differences as a function of children’s age, our findings showed that as the children’s age increased, the negative overlap between their avoidances and parents’ preferences (limitation area) decreased, suggesting that children tended to accept a larger number of food items among those particularly loved by their parents, probably because they were acquiring ‘‘more adult-like’’ tastes. Lastly, the child’s autonomous avoidances area tended to increase, suggesting that, through the exploration outside the parental repertory, older children were encountering new foodstuffs which they found unpleasant (but not which they loved, thus preferences seemed more stable). Four limitations of our research should be noted. First, the sample was not statistically representative of the Italian adolescent population and this prevented us from being able to generalize the observed results. Second, we asked participants to list only three preferences and only three avoidances. Probably, if participants had been free to mention as many food items as they wanted, it would have been possible to produce a more complete picture. The third weakness concerns the fact that we involved only
ARTICLE IN PRESS M. Guidetti, N. Cavazza / Appetite 50 (2008) 83–90
one parent in the sample, omitting the complexity of family influence processes. Finally, in this study, we could not consider the variables that might affect the relationship between the structure observed and the children’s age. On the basis of previous research, we can identify these potentially intervening factors. First, the period of reciprocal exposure increases with children’s age and therefore it should produce rising levels of resemblance. Second, children need to achieve autonomy from their parents and to prove their loyalty to their peers while at the same time they need to maintain family ties (Chapman & Maclean, 1993; Grotevant & Cooper, 1998): the former should lead to a decrease in parental influence, while the latter should lead to its stability. Third, a double motivation underlies the changes in adolescents’ eating habits (mainly at the level of attitudes, consumption, and meal patterns): on the one hand, the need to express their own freedom from parental control and, on the other, the implementation of likes and life styles considered proper for adults. While the former motivation should imply a decrease in similarities as children grow up, the latter should result in their increase. It is also necessary to bear in mind that family influences are multidirectional; therefore, as parental influence on children’s eating habits decreases, children’s influence on their parents could increase, for example. Lastly, the social and cultural pressures that exalt thinness (Chapman & Maclean, 1993) could induce adolescents to prefer healthier foods, moving them closer to parents’ likes. Despite these limitations, our research contributes an overview of the possible connections between parents’ and children’s preferences and avoidances, going beyond the approach based just on correlation analysis. Consistently with the previous research, our findings highlighted quite small overlapping areas between preferences and avoidances of the participant dyads’ members. However, our results also underlined the fact that the relationship between food repertories is not simply a matter of resemblance between ‘‘hedonic ratings’’; this relationship is a far more complex structure and it is important to separate preferences and avoidances, and to consider not only the reproduction, but also the intersection between opposite sign areas and the autonomy zones. It remains to be seen whether our descriptive model can be generalized and whether its structure will remain valid should a wider range of preferences and avoidances, and also of likes and dislikes be considered. Further research is needed regarding the development of the observed structure as children grow up, possibly through a longitudinal study design, and the role of intervening factors. In addition, the relation between the various processes of reciprocal influence and the structure of intersections observed should be studied. Finally, it would be very interesting to consider both parents (and not just one), as well as any siblings, in order to obtain a more complete model of the entire family.
89
The application value of the present data lies in the fact that increased knowledge in this domain may support nutritional education programs and the prevention of eating disorders, in particular, by facilitating the identification of the target of these interventions (children, parents, or families as a unit) and of the strategies to employ. Acknowledgement We wish to express our gratitude to Anna Rita Graziani and Alessandra Serpe for comments on earlier drafts of this manuscript. References Addessi, E., Galloway, A. T., Visalberghi, E., & Birch, L. L. (2005). Specific social influences on the acceptance of novel foods in 2–5-yearold children. Appetite, 45, 264–271. Birch, L. L. (1980). The relationship between children’s food preferences and those of their parents. Journal of Nutrition Education, 12, 14–18. Birch, L. L. (1999). Development of food preferences. Annual Review of Nutrition, 19, 41–62. Birch, L. L., & Fisher, J. O. (1998). Development of eating behaviors among children and adolescents. Pediatrics, 101, 539–549. Birch, L. L., Fisher, J. O., Grimm-Thomas, K., Markey, C. N., Sawyer, R., & Johnson, S. L. (2001). Confirmatory factor analysis of the child feeding questionnaire: A measure of parental attitudes, beliefs and practices about child feeding and obesity proneness. Appetite, 36, 201–210. Bisogni, C. A., Connors, M., Devine, C. M., & Sobal, J. (2002). Who we are and how we eat: A qualitative study of identities in food choice. Journal of Nutrition Education and Behaviour, 34, 128–139. Borah-Giddens, J., & Falciglia, G. A. (1993). A meta-analysis of the relationship in food preferences between parents and children. Journal of Nutrition Education, 25, 102–107. Breen, F. M., Plomin, R., & Wardle, J. (2006). Heritability of food preferences in young children. Physiology and Behavior, 88, 443–447. Brunstrom, J. M. (2004). Does dietary learning occur outside awareness? Consciousness and Cognition, 13, 453–470. Campbell, K. J., Crawford, D. A., & Ball, K. (2006). Family food environment and dietary behaviors likely to promote fatness in 5–6-year-old children. International Journal of Obesity, 30, 1272–1280. Carper, J. L., Orlet, F. J., & Birch, L. L. (2000). Young girls’ emerging dietary restraint and disinhibition are related to parental control in child feeding. Appetite, 35, 121–129. Chapman, G., & Maclean, H. (1993). ‘‘Junk food’’ and ‘‘healthy food’’: Meanings of food in adolescent women’s culture. Journal of Nutrition Education, 25, 108–113. Conner, M. (1993). Individualized measurement of attitudes towards food. Appetite, 20, 235–238. Conner, M., & Armitage, C. J. (2002). The social psychology of food. Buckingham: Open University Press. Cullen, K. W., Baranowski, T., Rittenberry, L., Cosart, C., Hebert, D., & de Moor, C. (2001). Child-reported family and peer influences on fruit, juice and vegetable consumption: Reliability and validity of measures. Health Education Research, 16, 187–200. De Bourdeaudhuij, I. (1996). Resemblance in health behaviours between family members. Archives of Public Health, 54, 251–273. De Bourdeaudhuij, I., & Van Oost, P. (2000). Personal and family determinants of dietary behaviour in adolescent and their parents. Psychology and Health, 15, 751–770. Erikson, E. H. (1963). Youth: Change and challenge. New York: Basic books. Fallon, A. E., & Rozin, P. (1983). The psychological bases of food rejections by humans. Ecology of Food and Nutrition, 13, 15–26.
ARTICLE IN PRESS 90
M. Guidetti, N. Cavazza / Appetite 50 (2008) 83–90
Feunekes, G. I. J., Stafleu, A., De Graaf, C., & Van Staveren, W. A. (1997). Family resemblance in fat intake in the Netherlands. European Journal of Clinical Nutrition, 51, 793–799. Fischler, C. (1990). L’homnivore. Paris: Odile Jacob. French, S. A., Story, M., Neumark-Sztainer, D., Fulkerson, J. A., & Hannan, P. (2001). Fast food restaurant use among adolescents: Associations with nutrient intake, food choices and behavioral and psychosocial variables. International Journal of Obesity, 25, 1823–1833. Grotevant, H. D., & Cooper, C. R. (1998). Individuality and connectedness in adolescent development: Review and prospects for research on identity, relationships, and context. Adolescence and society. In E. E. A. Skoe, & A. L. von der Lippe (Eds.), Personality development in adolescence: A cross national and life span perspective (pp. 3–37). Florence, KY: Taylor, & Frances/Routledge. Hannon, P. A., Bowen, D. J., Moinpour, C. M., & McLerran, D. F. (2003). Correlations in perceived food use between the family food preparer and their spouses and children. Appetite, 40, 77–83. Hertzler, A. A. (1983). Children’s food patterns—A review. II. Family and group behavior. Journal of the American Dietetic Association, 83, 555–560. Kratt, P., Reynolds, K., & Shewchuk, R. (2000). The role of availability as a moderator of family fruit and vegetable consumption. Health Education and Behaviour, 27, 471–482. Kremers, S. P., Brug, J., de Vries, H., & Engels, R. C. (2003). Parenting style and adolescent fruit consumption. Appetite, 41, 43–50. Lau, R. R., Quadrel, M. J., & Hartman, K. A. (1990). Development and change of young adults’ preventive health beliefs and behavior: Influence from parents and peers. Journal of Health and Social Behavior, 31, 240–259. Liem, D. G., Mars, M., & De Graaf, C. (2004). Sweet preferences and sugar consumption of 4- and 5-year-old children: Role of parents. Appetite, 43, 235–245. Logue, A. W., Logue, C. M., Uzzo, R. G., McCarty, M. J., & Smith, M. E. (1988). Food preferences in families. Appetite, 10, 169–180. Neumark-Sztainer, D., Hannan, P. J., Story, M., Croll, J., & Perry, C. (2003). Family meal patterns: Associations with sociodemographic characteristics and improved dietary intake among adolescents. Journal of the American Dietetic Association, 103(3), 317–322. Nicklas, T. A., Baranowski, T., Baranowski, J. C., Cullen, K., Rittenberry, L., & Olvera, N. (2001). Family and child-care provider influences on preschool children’s fruit, juice, and vegetable consumption. Nutrition Reviews, 59, 224–235. Patrick, H., & Nicklas, T. A. (2005). A review of family and social determinants of children’s eating patterns and diet quality. Journal of the American College of Nutrition, 24, 83–92. Patrick, H., Nicklas, T. A., Hughes, S. O., & Morales, M. (2005). The benefits of authoritative feeding style: Caregiver feeding styles and children’s food consumption patterns. Appetite, 44, 243–249. Patterson, T. L., Rupp, J. W., Sallis, I. F., Atkins, C. I., & Nader, P. R. (1988). Aggregation of dietary calories, fats, and sodium in Mexican–
American and Anglo families. American Journal of Preventive Medicine, 4, 75–82. Pliner, P. (1983). Family resemblances in food preferences. Journal of Nutrition Education, 15, 137–140. Pliner, P., & Pelchat, M. (1986). Similarities in food preferences between children and their siblings and parents. Appetite, 7, 333–342. Ritchey, N., & Olson, C. (1983). Relationship between family variables and children’s preference for and consumption of sweet foods. Ecology of Food and Nutrition, 13, 257–266. Rolls, B. J., Engell, D., & Birch, L. L. (2000). Serving portion size influences 5-year-old but not 3-year-old children’s food intakes. Journal of the American Dietetic Association, 100, 232–234. Rossow, I., & Rise, J. (1994). Concordance of parental and adolescent health behaviors. Social Science & Medicine, 38, 1299–1305. Rozin, P., & Fallon, A. E. (1980). The psychological categorization of foods and non-foods: A preliminary taxonomy of food rejections. Appetite, 1, 193–201. Rozin, P., & Fallon, A. E. (1987). A perspective on disgust. Psychological Review, 94, 23–41. Rozin, P. (1991). Family resemblance in food and other domains: The family paradox and the role of parental congruence. Appetite, 16, 93–102. Rozin, P., & Vollmecke, T. A. (1986). Food likes and dislikes. Annual Review of Nutrition, 6, 433–456. Rozin, P., Fallon, A., & Mandell, R. (1984). Family resemblance in attitudes to foods. Developmental Psychology, 20, 309–314. Skinner, J. D., Carruth, B. R., Wendy, B., & Ziegler, P. J. (2002). Children’s food preferences: A longitudinal analysis. Journal of the American Dietetic Association, 102, 1638–1647. Skinner, J., Carruth, B. R., Moran, J., Houck, K., Schmidhammer, J., Reed, A., et al. (1998). Toddlers’ food preferences: Concordance with family members’ preferences. Journal of Nutrition Education, 30, 17–22. Stafleu, A., Van Staveren, V. A., De Graaf, C., Burema, J., & Hautvast, J. G. A. J. (1995). Family resemblance in beliefs, attitudes and intentions towards consumption of 20 foods; a study among three generations of women. Appetite, 25, 201–216. Vartanian, L. R., Herman, C. P., & Polivy, J. (2007). Consumption stereotypes and impression management: How you are what you eat. Appetite, 48, 265–277. Vereecken, C. A., Keukelier, E., & Maes, L. (2004). Influence of mother’s educational level on food parenting practices and food habits of young children. Appetite, 43, 93–103. Videon, T. M., & Manning, C. K. (2003). Influences on adolescent eating patterns: The importance of family meals. Journal of Adolescent Health, 32, 365–373. Wardle, J., Cooke, L. J., Gibson, E. L., Sapochnik, M., Sheiman, A., & Lawson, M. (2003). Increasing children’s acceptance of vegetables; a randomized trial of parent-led exposure. Appetite, 40, 155–162. Weidner, G., Archer, S., Healy, B., & Matarazzo, J. D. (1985). Family consumption of low fat food: Stated preferences versus actual consumption. Journal of Applied Social Psychology, 15, 773–779.