Association between traditional food consumption and motives for food choice in six European countries

Appetite 53 (2009) 101–108

Contents lists available at ScienceDirect

Appetite journal homepage: www.elsevier.com/locate/appet

Research report

Association between traditional food consumption and motives for food choice in six European countries Zuzanna Pieniak a,*, Wim Verbeke a, Filiep Vanhonacker a, Luis Guerrero b, Margrethe Hersleth c a

Department of Agricultural Economics; Ghent University, Coupure links 653, 9000 Gent, Belgium IRTA–Monells, Finca Camps i Armet, E-17121 Monells, Spain c Nofima Food, Matforsk AS, Osloveien 1, 1430 A˚s, Norway b

A R T I C L E I N F O

A B S T R A C T

Article history: Received 25 September 2008 Received in revised form 13 May 2009 Accepted 24 May 2009

This study investigates the association between traditional food consumption and motives for food choice in six European countries. Cross-sectional data were collected through the TRUEFOOD panEuropean consumer survey (n = 4828) with samples representative for age, gender and region in Belgium, France, Italy, Norway, Poland and Spain. Importance attached to familiarity with a product is found to be strongly and positively associated with general attitude toward traditional food as well as traditional food consumption. The importance attached to convenience was negatively related to both general attitude toward traditional food and traditional food consumption, while the importance of weight control negatively influenced the general attitude. Natural content of food was positively associated with the attitude toward traditional food and traditional food consumption. The importance of price when purchasing food failed to be significantly related with general attitude and traditional food consumption both for the pooled sample as well as within each country except in Spain. The proposed model contributes to a better understanding of factors shaping the image and influencing the consumption of traditional foods in Europe. General attitude toward traditional foods, familiarity, and importance of food naturalness emerged as drivers for traditional food consumption. Importance attached to convenience and health acted as direct barriers to traditional food consumption, whereas importance of weight control emerged as an indirect barrier through lowering general attitude toward traditional foods. ß 2009 Elsevier Ltd. All rights reserved.

Keywords: Traditional food Consumer Motives Food Choice Questionnaire Structural equation modelling

Introduction In the last years there has been an increased interest in foods that are linked with a place or region of origin (Kuznesof, Tregear, & Moxey, 1997; Verbeke & Roosen, 2009). Consumers increasingly demand local food and foods with a traditional character or image, which are often perceived as higher quality (Chambers, Lobb, Butler, Harvey, & Traill, 2007; Fandos & Flavian, 2006) and more sustainable foods (Asebo, Jervell, Lieblein, Svennerud, & Francis, 2007; Risku-Norja, Hietala, Virtanen, Ketomaki, & Helenius, 2008) that fulfil a need for cultural identity and ethnocentrism (Chambers et al., 2007; Chryssochoidis, Krystallis, & Perreas, 2007; Vermeir & Verbeke, 2008). Traditional foods, which have

* Corresponding author. E-mail addresses: [email protected] (Z. Pieniak), [email protected] (W. Verbeke), fi[email protected] (F. Vanhonacker), [email protected] (L. Guerrero), margrethe.hersleth@nofima.no (M. Hersleth). 0195-6663/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.appet.2009.05.019

been defined as foods that are differentiated through particular qualitative aspects and have a specific cultural identity (Jordana, 2000), fit with the aforementioned evolution. Besides their growing appeal among consumers, traditional food products have received increasing policy support, in particular in the European Union. The EU Regulations 2081/92 and 2082/92 of 14 July 1992 have defined the standards for a designation under different collective trademarks: Protected Designation of Origin (PDO), Protected Geographical Indication (PGI), or Traditional Specialities Guaranteed (TSG) to a food product. These collective trademarks give producers the opportunity to promote their products and differentiate them on the market (Kuznesof et al., 1997). As a result of the policy and market interest, traditional foods have become increasingly attractive from an industry perspective, especially for small and medium sized enterprises. Emphasising product attributes generated by regional characteristics of the manufacturing area or by the use of traditional production practices creates new opportunities for marketers (Skuras & Vakrou, 2002). Moreover, cultivation of local raw materials and ingredients, which are mostly used in the production of traditional foods, contributes to a

102

Z. Pieniak et al. / Appetite 53 (2009) 101–108

Fig. 1. Hypothetical model linking motives for food choice with attitude and behaviour toward traditional foods.

more sustainable environment and employment of local people in rural areas (Trichopoulou, Vasilopoulou, Georga, Soukara, & Dilis, 2006) (Fig. 1). According to Bertozzi (1998) a traditional food product is a ‘‘representation of a group, it belongs to a defined space, and it is part of a culture that implies the cooperation of the individuals operating in that territory’’. Jordana (2000) derived from this sociological definition the following: ‘‘In order to be traditional, a product must be linked to a territory and it must also be part of a set of traditions, which will necessarily ensure its continuity over time’’. In 2006, the European Commission gave the following definition of ‘‘traditional’’ related to foods: ‘‘Traditional means proven usage in the Community market for a time period showing transmission between generations; this time period should be the one generally ascribed as one human generation, at least 25 years’’ (European Commission, 2006). Recently a definition of traditional food has been developed through the work of the EuroFIR Network of Excellence. This is an elaborative definition which includes statements about traditional ingredients, traditional composition and traditional type of production and/or processing (EuroFIR, 2007; Trichopoulou, Soukara, & Vasilopoulou, 2007). In Europe the only formal definition found for ‘‘traditional food products’’ (TFP) comes from the Italian Ministry of Agriculture, that defines TFP as ‘‘Agrifood products whose methods of processing, storage and ripening are consolidated with time according to uniform and constant local use’’ (Ministero Agricoltura, 1999). From the consumers’ perspective, Guerrero et al. (2009) defined a traditional food product as ‘‘a product frequently consumed or associated with specific celebrations and/or seasons, normally transmitted from one generation to another, made accurately in a specific way according to the gastronomic heritage, with little or no processing/ manipulation, distinguished and known because of its sensory properties and associated with a certain local area, region or country’’. Previous studies have focused on consumer preferences and behaviour toward ‘‘typical’’ products, such as Italian dry sausage (Conter et al., 2008), soppressata salami (Iaccarino, Di Monaco, Mincione, Cavella, & Masi, 2006), cheese (Kupiec & Revell, 1998) or honey (Sanzo, del Rio, Iglesias, & Vazquez, 2003). Furthermore, several studies have focused on consumer valuation of geographical indications and collective quality marks, which are often used on traditional food products. For example, van Ittersum, Meulenberg, van Trijp, & Candel (2007) concluded that consumers of regional products value regional certification labels. In contrast, Bonnet and Simioni (2001) concluded that it could not be taken for

granted that consumers in general value the quality signal provided by a PDO label on branded Camembert cheese. Apart from valuation studies focusing on specific food products with a traditional character, no research has been conducted to investigate consumers’ motives for purchasing traditional foods as a product category, or to explore why traditional foods are appealing to consumers. There is a large diversity in the production of traditional food in Europe, due to different climate conditions in the Northern and the Southern European countries. In general, two quite different cultural food systems have been profiled. First, the Northern European countries, where the range of foods traditionally produced was rather narrow; versus the Southern European countries, where this range and differentiation based on locality and tradition was rather wide (Jordana, 2000). According to Becker (2008) Southern European countries, such as Italy, Spain and France have a high number of collective quality marks, which can be regarded as possible candidates for a registration as a PDO or PGI. In contrast, Scandinavian and the Benelux countries have a very low number of collective quality marks. Hence, Southern European consumers are more likely to be confronted and to be familiar with traditional foods. Consequently, it is reasonable to observe and compare the factors influencing traditional food consumption across countries that differ in their cultural, geographical and gastronomic inheritance background, as has been demonstrated by, e.g. Askegaard and Madsen (1998). The purpose of this study is twofold. First, this paper assesses and tests the cross-cultural validity of eight motives for food choice based on the Food Choice Questionnaire (Steptoe, Pollard, & Wardle, 1995) in six European countries: Belgium, France, Italy, Norway, Poland and Spain. These motives are weight control, price, ethical concern, convenience, natural content, health, sensory appeal, and familiarity. The second and more specific objective of the study is to investigate the association of these motives for food choice with the specific choice for traditional foods and attitudes toward traditional foods. Materials and methods Research approach and sampling Quantitative descriptive data were collected through a crosssectional consumer survey with samples representative for age, gender and region in Belgium, France, Italy, Norway, Poland and Spain. The age range of the population was defined as 20–70 years. Total sample size was 4828 respondents, i.e. around 800 respondents in each of the six considered European countries. Participants were randomly selected from the representative TNS European Online Access Panel (Malhotra & Peterson, 2006) in line with the national population distributions with respect to age, gender and region. Such panels consist of individuals who have been recruited through off-line recruitment methods (e.g. random walk or street contact procedures) and who agreed to take part in future surveys. All contact and questionnaire administration procedures were electronic. Data collection was performed during the period from October 25 until November 9, 2007. Detailed socio-demographic characteristics of the national and pooled samples are provided in Table 1. Gender is equally distributed, which reflects that the population was intentionally not restricted to the main responsible person for food purchasing. Age distributions, mean age and mean household sizes match closely with the national census data of the respective countries. Table 1 also presents a proxy of socio-economic class, which was a subjective assessment of the household’s financial situation given on a 7-point interval scale ranging from ‘‘difficult’’ to ‘‘well off’’. Although this subjective measure may not necessarily reflect the

Z. Pieniak et al. / Appetite 53 (2009) 101–108

103

Table 1 Selected socio-demographic characteristics of the samples. Pooled sample (n = 4828)

Norway (n = 798)

Belgium (n = 826)

France (n = 801)

Spain (n = 800)

Italy (n = 800)

Poland (n = 803)

Gender (%) Female Male

49.2 50.8

49.1 50.9

49.4 50.6

51.9 48.1

47.4 52.6

47.3 52.7

50.2 49.8

Age (years) <35 35–55 >55

34.1 46.4 19.5

34.1 47.5 18.4

28.5 46.4 25.1

33.7 46.4 19.9

35.5 47.4 17.1

35.0 45.8 19.2

37.9 44.8 17.3

41.5 12.8

41.4 12.5

43.7 13.3

41.4 12.8

40.7 12.3

41.2 12.8

40.6 12.8

Household size (number) Mean 2.9 S.D. 1.3

2.6 1.3

2.7 1.3

2.7 1.2

3.1 1.3

3.2 1.3

3.0 1.4

24.8 31.5 43.7

17.8 28.6 53.6

35.5 32.5 32.0

18.9 36.2 44.9

29.8 32.8 37.4

21.3 31.0 47.7

Mean S.D.

Financial situation (%) Difficult–moderate Moderate Moderate–well off

24.6 32.1 43.3

actual financial situation of the households, its distribution reveals that different socio-economic classes are well represented in the sample. It should be noted that about half of the sample (52.6%) received education beyond the age of 18 years (high school or university), while 47.4% had a lower level of education (primary or secondary school education). The sample is herewith slightly biased towards higher education, which may be attributed to the use of an electronic survey method. Questionnaire content and pre-testing Participants were asked to complete a self-administered structured electronic questionnaire. The questionnaire consisted of five sections: (1) behavioural and attitudinal items relating to food purchasing in general, (2) items probing for consumers’ definition, attitude and perceptions of traditional foods, (3) consumer awareness of and associations with quality and origin labels, (4) items probing for consumer’s definition and acceptance of innovations in traditional food products, and (5) personal data relating to lifestyle and general interests, and socio-demographics. The questionnaire development was based on the output of a comprehensive qualitative research task about traditional food performed in the six named countries (Guerrero et al., 2009). The focus of this paper is on assessing the association of motives relating to food purchasing in general (as covered in the first section of the questionnaire) with traditional food consumption. The master questionnaire was developed in English and translated in the national languages using the procedure of back-translation (Brislin, 1970; Maneesriwongul & Dixon, 2004). Following back-translation, the questionnaire was extensively pretested by the researchers through personal interviews with 15–20 respondents in each of the countries in order to identify and eliminate potential problems and to ensure linguistic equivalence. Fieldwork started after editing, correcting, electronic programming and additional pre-testing of the electronic version of the questionnaire. The average time for completing the total questionnaire ranged from 290 3300 in France to 330 3600 in Poland. Measurement and scaling Potential motives for choosing food were almost entirely based on the Food Choice Questionnaire (FCQ) as described by Steptoe et al. (1995). Only the most appropriate and relevant items for the case of traditional food were included. The selection of items was based on findings from exploratory focus group discussions

(Guerrero et al., 2009) and literature review. For example, the mood FCQ dimension was not included in the questionnaire. As a result, eight dimensions were included, namely weight control, price, ethical concern, convenience, natural content, health, sensory appeal, and familiarity. Respondents were asked to endorse the statement ‘‘It is important to me that food I eat on a typical weekday . . .’’ (see Table 3) for each of the 24 items (three items per dimension) by scoring on a 7-point scale ranging from ‘‘not important at all’’ (1) to ‘‘extremely important’’ (7). General attitude toward traditional food was measured using three items that were scored on 7-point semantic differential scales. Respondents were presented with the statement ‘‘Please indicate which word best describes how you feel when you eat traditional food’’. The bipolar adjectives were unhappy/happy, dull/exiting, and terrible/delightful. These items have previously been used to assess general attitude in both marketing (e.g. Stayman & Batra, 1991) and food consumption behaviour studies (e.g. Sparks & Guthrie, 1998). Traditional food consumption was a self-reported single item measure: ‘‘To what extent do you consider yourself a consumer of traditional food?’’ on a 7-point Likert scale ranging from ‘‘very much’’ (7) to ‘‘not at all’’ (1). Statistical analyses Completed questionnaires were edited by the field research agency in order to ensure accuracy and precision of the response prior to coding and transcription of the data in SPSS 15.0 format. Given the large sample sizes and very low numbers of missing responses, pairwise deletion was used as the method for treating missing values. In line with the first objective of our study, six multi-group confirmatory factor analysis models were estimated following the procedure recommended by Steenkamp and Baumgartner (1998), and using the robust maximum likelihood procedure in LISREL 8.72 in a stepwise approach, with each country representing one group. Next, a maximum likelihood confirmatory factor analysis on the pooled sample (exploring discriminant and convergent validity) was performed. In order to meet the second objective of the study, structural equations model parameters have been estimated and the general fit of the model has been assessed by means of LISREL 8.72. With the use of structural equation modelling (SEM), the examination of all the relationships between constructs and items is performed simultaneously, which is a substantial advantage compared with single equation modelling (Bollen, 1989). Due to the large sample size the x2 may not be the

Z. Pieniak et al. / Appetite 53 (2009) 101–108

104 Table 2 Results from measurement invariance analysis. Model

Constraints added

Satorra–Bentler x2

df

RMSEA

NNFI

CFI

Dx2

Ddf

Configural Metric invariance Scalar invariance Factor variance invariance Error variance invariance Identity

Baseline—no constraints Factor loadings Item intercepts Factor covariance and factor variance Measurement error Latent factor means

7279.15 7836.01 10368.78 12029.23 14206.65 16454.3

1842 1932 2017 2292 2427 2477

0.061 0.062 0.072 0.073 0.077 0.084

0.975 0.974 0.965 0.965 0.959 0.94

0.98 0.978 0.969 0.964 0.957 0.949

556.86 2532.77 1660.45 2177.42 2247.65

90 85 275 135 50

most appropriate measure of goodness-of-fit (Browne & Cudeck, 1993). Therefore, three other indices will be reported: the Root Mean Square Error of Approximation (RMSEA), the goodness-of-fit index (GFI) and the Comparative Fit Index (CFI). Values below 0.08 for RMSEA (Browne & Cudeck, 1993) and above 0.90 for GFI and CFI (Bollen, 1989) suggest an acceptable fit of the model.

Scholderer, Grunert, & Brunsø, 2005) suggest that scalar invariance suffices for pooling the data. Following the latter studies, and considering that the identity model is not very clearly rejected by the data, pooling of data is still an option.

Measurement invariance

Confirmatory factor analysis was performed to determine whether the construct measures actually converge the intended latent variable or share a high proportion of variance in common (convergent validity) and whether the constructs are distinct from each other (discriminant validity). In order to assess convergent validity, the procedures suggested by Hair, Black, Babin, Anderson, & Tatham (2006) have been followed. Standardised factor loadings, reliability estimates and the variance extracted are presented in Table 3. The individual item loadings on the constructs were all highly significant with values ranging from 0.68 to 0.87. No cross loadings worth mentioning appeared. Hence, all the items were considered in the interpretation of the factors (Hair et al., 2006). Additionally, Cronbach’s alpha internal reliability coefficients were above the threshold value of 0.7 for satisfactory scales. Finally, in our study, the variance extracted measures exceeded the 0.5 threshold (Fornell & Larcker, 1981). Our results satisfy the three criteria for convergent validity for the internal constructs (Bagozzi, Li, & Phillips, 1991; Hair et al., 2006). Intercorrelations between the dimensions are presented in Table 4. All correlation coefficients were significant and below 0.80, thus (severe) multicollinearity is not a concern in the present data (Tabachnick & Fidell, 2001). Although three of the correlations are above 0.7, the relatively high composite reliability measures (>0.7), as well as our large sample (n = 4828) can effectively protect the data against eventual deleterious effects of multicollinearity (Grewal, Cote, & Baumgartner, 2004), unless the multicollinearity would have been severe (i.e. correlations above 0.8). To further assess the discriminant validity of the subset of measures, we adopted the procedure recommended by Hair et al. (2006) based on Fornell and Larcker (1981). All variance-extracted estimates displayed in Table 3 were greater than the corresponding interconstruct squared correlation estimates in Table 4. Therefore, these tests satisfy the criteria for discriminant validity.

The present data were collected in six different European countries, which implies some concerns related to the crosscultural validity of the collected information. In order to test whether the psychometric characteristics of the measures are stable across the six countries a measurement invariance analysis was conducted through a stepwise multi-group confirmatory factor analysis (Steenkamp & Baumgartner, 1998). Each step constrained a particular matrix to be equal across all groups. Each restricted model has been nested within a less restricted one. This procedure allows comparing models statistically using the difference in the x2 statistics, degrees of freedom, as well as other goodness-of-fit parameters (see Table 2). Although the samples involved in the analysis were so large that the Satorra–Bentler x2 was significant at every step, the RMSEA values indicated acceptable fit for the first five models reported in Table 2 and a tentatively acceptable fit for the identity model (RMSEA = 0.084, NNFI = 0.94, CFI = 0.95). These findings indicate that the psychometric characteristics of the measures were invariant across the six-country specific samples. Specifically, configural invariance implies that participants belonging to different groups, or countries in this particular study, conceptualise constructs in the same way; thus that the same underlying constructs are measured in all involved countries. Metric invariance has also been supported, meaning that the observed variables (i.e. motives and barriers for traditional food consumption) are measured according to the same scale units. Thus, there is conceptual agreement in terms of the type and the number of underlying constructs as well as the degree to which items are associated with each construct. Furthermore, the measurement model showed scalar invariance, meaning that the observed variables are measured on common interval scales across groups, i.e. according to the same scale units and scale locations. Factor covariance invariance indicates that the constructs measured using the FCQ items have the same interrelationships across countries. Factor variance invariance indicates that the constructs measured by the FCQ have the same variability across countries. Error variance invariance indicates that item reliabilities are the same across the samples. Finally, identity implies that the means of the underlying constructs are the same across the samples. Rungtusanatham, Ng, Zhao, & Lee (2008) as well as Steenkamp and Baumgartner (1998) recommended pooling data only if all sub-dimensions of measurement invariance are satisfied for all measures. In this study, the identity model suggests only a tentatively acceptable fit. The RMSEA is close to acceptable fit below 0.08, and the other fit statistics are acceptable. However, other studies (Olsen, Scholderer, Brunso, & Verbeke, 2007;

Confirmatory factor analysis and construct validity

Model validation The next step in the analysis concerns testing the structural model that relates each of the motives for food choice to the overall attitude toward traditional foods and behaviour toward traditional food. The presumed model performed well on the pooled sample (Table 5). The x2 for the model was 5245.31 with 306 degrees of freedom (p < 0.001); the RMSEA value was 0.058; the GFI was 0.91 and the CFI was 0.98. Also the other goodness-of-fit indices were satisfactory. Importance of weight control and convenience were negatively associated with consumers’ general attitude toward traditional food. However, those associations were relatively weak. Hence,

Z. Pieniak et al. / Appetite 53 (2009) 101–108

105

Table 3 Factor loadings, reliability estimates and variance extracted for motives for food choice. Constructs and items It is important to me that the food I eat on a normal weekday . . .

Standardised factor loading

Weight control Helps me control my weight Is low in fat Is low in calories

0.78 0.84 0.77

Price Is not expensive Is good value for money Is cheap

0.83 0.68 0.77

Ethical concern Is produced using ethical production methods (e.g. sustainable, animal friendly, without child labour, etc) Is produced/packaged in an environmentally friendly way Supports the local economy

Composite reliability

Variance extracted

0.90

0.64

0.79

0.58

0.84

0.57

0.84

0.64

0.85

0.65

0.89

0.74

0.89

0.69

0.74

0.54

0.78 0.77 0.71

Convenience Is easy to plan, buy and prepare Takes very little time to prepare Can be cooked very easily

0.82 0.78 0.80

Natural content Contains natural ingredients Contains no artificial ingredients Contains no additives

0.87 0.78 0.77

Health Contains a lot of vitamins and minerals Keeps me healthy Is nutritious

0.87 0.87 0.81

Sensory appeal Smells nice Has a pleasant texture Tastes well

0.86 0.82 0.82

Familiarity Is familiar Is what I usually eat Is like the food I ate when I was a child

0.77 0.76 0.68

All factor loadings are significant at p < 0.001. Fit statistics for the pooled data RMSEA = 0.058; GFI = 0.91; CFI = 0.98.

people who attached more importance to control their weight and who attached more importance to convenience in food purchasing and consumption had less positive attitudes toward traditional food. This indicates that traditional foods failed to appeal to consumers in terms of convenience and weight control. On the contrary, familiarity was positively associated with general attitude toward traditional food. Additionally, importance of convenience and health were found to have a negative direct relationship with the claimed consumption of traditional food. On the contrary, importance of natural content and familiarity when purchasing food were positively associated with traditional food consumption. The structural equation analysis supports that attitudes toward traditional food have a direct positive and relatively strong effect on traditional food consumption, which corroborates classical attitude–behaviour theories.

In order to provide insights in cross-country differences in the association between food choice motives and traditional food consumption, a multi-group structural equation modelling analysis has been performed. In general, the model performed well x2 = 7771.35, df = 1926; p < 0.001; RMSEA = 0.061; NNFI = 0.98; CFI = 0.98 (Table 5). The results of the structural equation analysis showed that importance of weight control had a significant negative association with general attitude toward traditional food in Norway and Poland, though no significant direct relationship with consumption in any of the six countries. Price had no significant relationship with general attitude in all countries, except in Spain where the relationship was negative; and no significant association with traditional food consumption in all investigated countries. Hence, although Spanish consumers with a stronger importance of price

Table 4 Correlation matrix of the ‘‘motives for food choice’’ dimensions. Construct

1

2

3

4

5

6

7

8

1. 2. 3. 4. 5. 6. 7. 8.

1.00 0.46 0.53 0.46 0.55 0.61 0.42 0.43

1.00 0.49 0.60 0.48 0.57 0.57 0.47

1.00 0.41 0.71 0.67 0.51 0.51

1.00 0.42 0.50 0.53 0.46

1.00 0.74 0.58 0.53

1.00 0.76 0.55

1.00 0.48

1.00

Weight control Price Ethical concern Convenience Natural content Health Sensory appeal Familiarity

Note: All correlations are statistically significant at p < 0.01 (two-tailed).

Z. Pieniak et al. / Appetite 53 (2009) 101–108

106

Table 5 Structural model (standardised solution) per country and for the total pooled sample. Construct

Path

Construct

Belgium

France

Italy

Norway

Poland

Spain

Total

Weight control Price Ethical concern Convenience Natural content Healthiness Sensory appeal Familiarity Weight control Price Ethical concern Convenience Natural content Healthiness Sensory appeal Familiarity General attitude

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

General attitude General attitude General attitude General attitude General attitude General attitude General attitude General attitude Traditional food consumption Traditional food consumption Traditional food consumption Traditional food consumption Traditional food consumption Traditional food consumption Traditional food consumption Traditional food consumption Traditional food consumption

ns ns ns 0.19 ns ns 0.62 0.31 ns ns ns ns ns ns ns ns 0.34

ns ns ns 0.15 ns ns 0.30 0.39 ns ns ns 0.12 0.46 0.36 ns 0.30 0.19

ns ns ns 0.20 ns ns ns 0.39 ns ns ns ns ns ns ns 0.35 0.33

0.32 ns ns ns ns ns ns 0.24 ns ns ns 0.25 ns ns ns 0.39 0.53

0.32 ns ns ns ns ns ns 0.32 ns ns ns ns ns ns ns 0.30 0.53

ns 0.19 ns ns ns ns ns 0.24 ns ns ns 0.15 ns ns ns 0.19 0.45

0.07 ns ns 0.08 ns ns ns 0.20 ns ns ns 0.17 0.25 0.31 ns 0.37 0.57

Goodness-of-fit statistics for the multi-group analysis: x2 (1926) = 7771.35, p < 0.001; RMSEA = 0.061; NNFI = 0.98; CFI = 0.98. Goodness-of-fit statistics for the total sample: x2 (306) = 5245.31, p < 0.001; RMSEA = 0.058; NNFI = 0.98; CFI = 0.98.

felt less positive about traditional foods, their price consciousness did not deter directly from traditional food consumption. Ethical concern neither had a significant association with general attitude nor with consumption of traditional food in all six European counties. Convenience had a negative association with general attitude toward traditional foods in Belgium, France, and Italy, and a negative relation with traditional food consumption in France, Norway and Spain. Importance of natural content and health had no significant association with general attitude and consumption in all investigated counties, except in France, where importance of natural content was found to be positively and relatively strong related to traditional foods consumption, whilst importance of health was negatively and moderately associated with consumption. Sensory appeal was significantly and positively associated with general attitude toward traditional food in Belgium and France. Interestingly, in Belgium this association was relatively strong as compared to the other countries (0.62). The association between sensory appeal and consumption was not significant in all investigated countries. A moderate to relatively strong association between familiarity and general attitude toward traditional food was confirmed in all countries. Additionally, familiarity was confirmed to have a direct moderate and positive relationship with traditional food consumption in almost all countries, except for Belgium. Furthermore, the relatively strong relationship between general attitude and traditional food consumption was confirmed in all six investigated countries. Discussion and conclusions The first objective of this study was to assess and test crosscultural validity of eight motives for food choice in six European countries: Belgium, France, Italy, Norway, Poland and Spain. The results indicate that the considered dimensions, i.e. weight control, price, ethical concern, convenience, natural content, health, sensory appeal, and familiarity, measured with the three items each as detailed in Table 3, have a similar meaning and similar structural characteristics across cultures within Europe. Those constructs are proven to be cross-culturally valid, which means that observed differences are real differences, in the sense that they do not stem from cross-cultural differences in the interpretation of items or concepts. These findings contradict those obtained by Eertmans, Victoir, Notelaers, Vansant, & Van den Bergh (2006), who suggested that the FCQ items and the underlying constructs might have different connotations in different European countries. One of the explanations could be that not all of the FCQ items

(Steptoe et al., 1995) were included in the present study (e.g. mood was not included). In addition we have used only three items describing each dimension. Among those items, some were slightly adapted based on the results of a previous qualitative exploratory study that was performed in the same countries as covered in the survey discussed in this paper (Guerrero et al., 2009). In our study, the set of items related to the underlying constructs were crossnationally congeneric, so the motive dimensions can be conceptualised in the same way across the countries involved in this study, i.e. Belgium, France, Italy, Norway, Poland and Spain. Next, configural, metric and scalar forms of invariance were established, thus comparison of means between countries might be performed and is meaningful since the measurement scales are similar across the involved countries. Additionally, configural invariance, metric invariance, factor variance invariance and error variance invariance were satisfied, allowing comparison of correlations between scale scores between the six investigated countries (Steenkamp & Baumgartner, 1998). Furthermore, the original motive dimensions used in this study were proven to display convergent and discriminant validity. The second major objective of this study was to assess the impact of the motives for food choice on traditional food consumption. By using structural equation modelling, we were able to estimate the strength and direction of direct and indirect relationships between the different constructs on one hand, and consumers’ attitudes toward traditional food and traditional food consumption on the other hand. The study herewith sheds light on why traditional foods appeal to consumers or not in different European countries in relation to their motives for purchasing food in general. Importance of weight control had a significant negative association with general attitude toward traditional food in Norway and Poland, where traditional foods were mostly recognised as rather fatty, as was concluded also from the exploratory focus group study (Guerrero et al., 2009). Furthermore, traditional foods have been already associated with ‘special occasions’ in Norway (Guerrero et al., in press) and specialty dishes that are consumed on festive occasions. This kind of food eaten on special occasions may be perceived as more sophisticated, more ‘hedonic’ and providing primarily pleasure rather than nutrition and health benefits, and therefore less agreeing with controlling weight. In general, price is recognised as an important factor influencing food choice in general (Furst, Connors, Bisogni, Sobal, & Falk, 1996; Jomori, Proenca, & Calvo, 2008; Steptoe et al., 1995).

Z. Pieniak et al. / Appetite 53 (2009) 101–108

Furthermore, based on qualitative focus group discussions performed in the UK, Chambers et al. (2007) suggested price together with inconvenience as barriers preventing consumers from buying more local products. However, in our study, the importance of price when buying daily meals failed to be significantly associated with both general attitude toward traditional food and traditional food consumption both for the pooled sample as well as within each country, except in Spain. Apparently, European consumers are aware of the potential price premiums associated with traditional foods relative to other foods. This does neither directly deter from traditional food consumption, nor does it negatively influence their attitude toward traditional foods. The difference as compared to the conclusions by Chambers et al. (2007) might be attributed either to the different cultural setting and research method, though it may also suggest that the concept of traditional foods (and related price perceptions) is not simply equivalent to the concept of local food. This is in line with the consumer driven definition set forth by Guerrero et al. (2009), which suggests that ‘traditional’ means more than just ‘local’. Another explanation could be that the respondents in our sample were relatively highly educated (thus possibly also earning more money) and hence price would be for them less relevant for product choice. Ethical concern has not appeared to have a significant relation with general attitudes toward traditional food and traditional food consumption. This finding indicates that ethical issues related, e.g. to the production and image of traditional foods neither act as a motive nor as a barrier to feel favourable about traditional foods or to purchase and consume traditional foods more frequently. Furthermore, importance attached to convenience revealed to be negatively related to both general attitude and traditional food consumption, which corroborates Chambers et al. (2007) for their case of local foods in the UK. People who were more focused on convenience in food choice had a more negative attitude and indicated a lower consumption of traditional food products. This finding reveals that traditional foods fail to appeal to consumers in terms of perceived convenience. In contrast, importance of natural content was positively associated with traditional food consumption. This association was particularly strong in France. Thus, the more important natural content was for consumers related to food in general, the higher their traditional food consumption. The term ‘natural’ has been already associated with other emotional terms such as traditional and homemade (Bender, 1989). Additionally, Kuznesof et al. (1997) showed that regional food was perceived as ‘‘authentic’’ and as occurring naturally or abundantly. These findings suggest that the natural character or image of traditional foods constitutes one of its major assets vis-a`-vis consumers. In general, health is recognised as one of the most important determinants of food choice (Verbeke, 2008). Traditional foods may have potential health characteristics (Trichopoulou et al., 2007), which have been tested over time, e.g. in the case of the Mediterranean diet (de Lorgeril & Salen, 2006; Trichopoulou, 2007). However, differences in dietary patterns between European populations living in different geographical regions are quite substantial and key elements of those eating habit and nutritional differences are related to traditional foods (Slimani et al., 2002). In our study, a negative association between health as a motive for food choice and traditional food consumption has been found, particularly in France. This finding suggests that traditional foods are not perceived as fitting with a healthy food choice, despite their perceived natural character. Potential explanations may stem from the perception that traditional foods are often full fat and energy dense products, or products that underwent only minimal conservation processing or packaging, and hence are perhaps perceived as entailing higher microbial risk.

107

Next, the results indicated that in general sensory appeal and perceived importance of sensory experiences, including smell, taste and appearance, were not significantly associated with general attitude and behaviour toward traditional foods. Only in Belgium and France higher interest in sensory performance of food made traditional foods more appealing for consumers (as reflected in the positive path toward attitude). However, sensory performance did not directly influence the choice for traditional foods. Importance attached to familiarity with a product was shown to positively influence traditional food consumption behaviour, which indicates that people who attached more importance to familiar products are more likely to opt for a traditional food product. This finding corroborates one of the important elements of the definition of traditional foods, namely the fact that it concerns products that are frequently consumed and whose consumption habit is transferred from one generation to the other (Guerrero et al., 2009). Additionally, our results confirm a positive relationship between attitude and behaviour, which is in line with classical attitude–behaviour theories (Ajzen, 1991; Ajzen & Fishbein, 1980). In this study, general attitude toward traditional food had the highest path coefficient linking with traditional food consumption. Some limitations of this study should be noted. Our study focused on a single item and self-reported measure of traditional food consumption behaviour. Although using a single item measure for behaviour is a common practice, the self-reported and subjective nature of the measure is more problematic. Specifically, it might have been subjected to social desirability, post-rationalisation, and cognitive dissonance or consonance and hence may deviate from actual behaviour. Therefore, further validation studies including objective measures of market place behaviour are recommended. Furthermore, this paper focuses on traditional food as a concept, which is quite broad and open to multiple interpretations, therefore leaving the question whether similar findings apply for different types of traditional foods (e.g. typical everyday food products versus the specialities type products). It should be stressed also that consumers’ (purchasing) behaviour is likely to be conditioned by more variables than the ones dealt with in this study (among others, environmental factors, information variables, socio-demographic characteristics, emotional aspects). Further research investigating the impact of a broader array of factors together with those investigated in this study is recommended. Finally, our sample is biased toward higher educated Europeans which may be associated with the online data collection method that requires higher computer literacy and web access. Nevertheless, the proportions of education level categories are consistent across countries. Traditional foods are closely connected to the culture and the identity of the population where they are produced; they carry a strong symbolic value (Guerrero et al., 2009). Despite the obvious cultural differences between the six considered countries, the results obtained from a multi-group analysis were very similar. General attitude toward traditional foods, familiarity, and importance of food naturalness emerged as drivers for traditional food consumption, whereas importance of convenience and health emerged as barriers to traditional food consumption.

References Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behaviour. Englewood Cliffs, NJ: Prentice-Hall. Asebo, K., Jervell, A. M., Lieblein, G., Svennerud, M., & Francis, C. (2007). Farmer and consumer attitudes at farmers markets in Norway. Journal of Sustainable Agriculture, 30, 67–93.

108

Z. Pieniak et al. / Appetite 53 (2009) 101–108

Askegaard, S., & Madsen, T. K. (1998). The local and the global: exploring traits of homogeneity and heterogeneity in European food cultures. International Business Review, 7, 549–568. Bagozzi, R. P., Li, Y. J., & Phillips, L. W. (1991). Assessing construct-validity in organizational research. Administrative Science Quarterly, 36, 421–458. Becker, T. (2008). European food quality policy: the importance of geographical indications, organic certification and food quality insurance schemes in European countries. Estey Centre Journal of Internal Law and Trade Policy, 10, 111–130. Bender, A. E. (1989). What is natural. Food Chemistry, 33(1), 43–51. Bertozzi, L. (1998). Tipicidad alimentaria y dieta mediterra´nea. In A. Medina, F. Medina, & G. Colesanti (Eds.), El color de la alimentacio´n mediterra´nea. Elementos sensoriales y culturales de la nutricio´n (pp. 15–41). Barcelona: Icaria. Bollen, K. A. (1989). Structural equations with latent variables. New York: Wiley. Bonnet, C., & Simioni, M. (2001). Assessing consumer response to protected designation of origin labelling: a mixed multinomial logit approach. European Review of Agricultural Economics, 28, 433–449. Brislin, R. W. (1970). Back-translation for cross-cultural research. Journal of Crosscultural Psychology, 1, 185–216. Browne, M. & Cudeck, R. (1993). Alternative ways of assessing model fit. In Bollen, K. & Long, J. (Eds.), Testing structural equation models. Newbury Park: CA: Sage. 136– 162. Chambers, S., Lobb, A., Butler, L., Harvey, K., & Traill, B. (2007). Local, national and imported foods: a qualitative study. Appetite, 49, 208–213. Chryssochoidis, G., Krystallis, A., & Perreas, P. (2007). Ethnocentric-beliefs and countryof-origin (COO) effect—impact of country, product and product attributes on Greek consumers’ evaluation of food products. European Journal of Marketing, 41, 1518– 1544. Conter, M., Zanardi, E., Ghidini, S., Pennisi, L., Vergara, A., Campanini, G., et al. (2008). Consumers’ behaviour toward typical Italian dry sausages. Food Control, 19, 609– 615. de Lorgeril, M., & Salen, P. (2006). The Mediterranean-style diet for the prevention of cardiovascular diseases. Public Health Nutrition, 9, 118–123. Eertmans, A., Victoir, A., Notelaers, G., Vansant, G., & Van den Bergh, O. (2006). The Food Choice Questionnaire: factorial invariant over western urban populations? Food Quality and Preference, 17, 344–352. EuroFIR (2007). FOOD-CT-2005-513944. EU6th framework food quality and safety programme, www.eurofir.net. European Commission (2006). Council Regulation (EC) No 509/2006 of 20 March 2006 on agricultural products and foodstuffs as traditional specialties guaranteed. Official Journal of the European Union L 93/1. Fandos, C., & Flavian, C. (2006). Intrinsic and extrinsic quality attributes, loyalty and buying intention: an analysis for a PDO product. British Food Journal, 108, 646–662. Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 39– 50. Furst, T., Connors, M., Bisogni, C. A., Sobal, J., & Falk, L. W. (1996). Food choice: a conceptual model of the process. Appetite, 26, 247–265. Grewal, R., Cote, J. A., & Baumgartner, H. (2004). Multicollinearity and measurement error in structural equation models: implications for theory testing. Marketing Science, 23, 519–529. Guerrero L., Claret A., Verbeke W., Enderli E., Zakowska-Biemans S., Vanhonacker F., et al. (in press). Perception of traditional food products in six European regions using free word association. Food Quality and Preferences, doi:10.1016/j.foodqual.2009.06.003. Guerrero, L., Gua`rdia, M. D., Xicola, J., Verbeke, W., Vanhonacker, F., Zakowska, S., et al. (2009). Consumer-driven definition of traditional food products and innovation in traditional foods. A qualitative cross-cultural study. Appetite, 52, 345–354. Hair, J., Black, W., Babin, B., Anderson, R., & Tatham, R. (2006). Multivariate data analysis. New Jersey: Pearson Education Inc. Iaccarino, T., Di Monaco, R., Mincione, A., Cavella, S., & Masi, P. (2006). Influence of information on origin and technology on the consumer response: the case of soppressata salami. Food Quality and Preference, 17, 76–84.

Jomori, M. M., Proenca, R. P. D., & Calvo, M. C. M. (2008). Food choice factors. Revista De Nutricao-Brazilian Journal of Nutrition, 21, 63–73. Jordana, J. (2000). Traditional foods: challenges facing the European food industry. Food Research International, 33, 147–152. Kupiec, B., & Revell, B. (1998). Speciality and artisanal cheeses today: the product and the consumer. British Food Journal, 100, 236–243. Kuznesof, S., Tregear, A., & Moxey, A. (1997). Regional foods: a consumer perspective. British Food Journal, 99, 199–206. Malhotra, N. K., & Peterson, M. (2006). Basic marketing research: a decision-making approach. Upper Saddle River: Pearson Education/Prentice Hall. Maneesriwongul, W., & Dixon, J. K. (2004). Instrument translation process: a methods review. Journal of Advanced Nursing, 48, 175–186. Ministero Agricoltura (1999). Decreto Legislativo 30 Aprile 1998 n. 173. Decreto Ministero Agricoltura 8 settembre 1999 n. 350. Olsen, S. O., Scholderer, J., Brunso, K., & Verbeke, W. (2007). Exploring the relationship between convenience and fish consumption: a cross-cultural study. Appetite, 49, 84–91. Risku-Norja, H., Hietala, R., Virtanen, H., Ketomaki, H., & Helenius, J. (2008). Localisation of primary food production in Finland: production potential and environmental impacts of food consumption patterns. Agricultural and Food Science, 17, 127–145. Rungtusanatham, M., Ng, C. H., Zhao, X., & Lee, T. S. (2008). Pooling data across transparently different groups of key informants: measurement equivalence and survey research. Decision Sciences, 39, 115–145. Sanzo, M. J., del Rio, A. B., Iglesias, V., & Vazquez, R. (2003). Attitude and satisfaction in a traditional food product. British Food Journal, 105, 771–790. Scholderer, J., Grunert, K. G., & Brunsø, K. (2005). A procedure for eliminating additive bias from cross-cultural survey data. Journal of Business Research, 58, 72–78. Skuras, D., & Vakrou, A. (2002). Consumers’ willingness to pay for origin labelled wine. A Greek case study. British Food Journal, 104, 898–912. Slimani, N., Fahey, M., Welch, A., Wirfa¨lt, E., Stripp, C., Bergstro¨m, E., et al. (2002). Diversity of dietary patterns observed in the European Prospective Investigation into Cancer and Nutrition (EPIC) project. Public Health Nutrition, 5, 1311–1328. Sparks, P., & Guthrie, C. A. (1998). Self-identity and the theory of planned behavior: a useful addition or an unhelpful artifice? Journal of Applied Social Psychology, 28, 1393–1410. Stayman, D. M., & Batra, R. (1991). Encoding and retrieval of Ad affect in memory. Journal of Marketing Research, 28, 232–239. Steenkamp, J. B. E. M. , & Baumgartner, H. (1998). Assessing measurement invariance in cross-national consumer research. Journal of Consumer Research, 25, 78–90. Steptoe, A., Pollard, T. M., & Wardle, J. (1995). Development of a measure of the motives underlying the selection of food—The Food Choice Questionnaire. Appetite, 25, 267–284. Tabachnick, B. G., & Fidell, L. S. (2001). Using Multivariate Statistics. Needham Heights, MA: Allyn & Bacon. Trichopoulou, A. (2007). Mediterranean diet, traditional foods, and health: evidence from the Greek EPIC cohort. Food and Nutrition Bulletin, 28, 236–240. Trichopoulou, A., Soukara, S., & Vasilopoulou, E. (2007). Traditional foods: a science and society perspective. Trends in Food Science & Technology, 18, 420–427. Trichopoulou, A., Vasilopoulou, E., Georga, K., Soukara, S., & Dilis, V. (2006). Traditional foods: why and how to sustain them. Trends in Food Science & Technology, 17, 498– 504. van Ittersum, K., Meulenberg, M. T. G., van Trijp, H. C. M., & Candel, M. J. J. M. (2007). Consumers’ appreciation of regional certification labels: a pan-European study. Journal of Agricultural Economics, 58, 1–23. Verbeke, W. (2008). Impact of communication on consumers’ food choices. Proceedings of the Nutrition Society, 67, 281–288. Verbeke, W., & Roosen, J. (2009). Market differentiation potential of origin, quality and traceability labelling. Estey Centre Journal of International Law and Trade Policy, 10, 20–35. Vermeir, I., & Verbeke, W. (2008). Sustainable food consumption among young adults in Belgium: theory of planned behaviour and the role of confidence and values. Ecological Economics, 64, 542–553.