The Application of the Theory of Planned Behaviour to Consumer Food Choice

ABSTRACT - Human food choice is a complex phenomenon influenced by a wide range of factors. The theory of planned behaviour offers a means for trying to understand the roles of some of these factors. Evidence will be presented for the inclusion of measures of self-identity and moral obligation. Problems with the measurement of perceived control and possible means for overcoming them will also be discussed. Future research will include the application of the causal modelling technique LISREL and collaborative work with Aarhus and Helsinki seeking to establish the cross-cultural validity of such approaches.

Citation:

R. Shepherd, P. Sparks, and C.A. Guthrie (1995) ,"The Application of the Theory of Planned Behaviour to Consumer Food Choice", in E - European Advances in Consumer Research Volume 2, eds. Flemming Hansen, Provo, UT : Association for Consumer Research, Pages: 360-365.

European Advances in Consumer Research Volume 2, 1995      Pages 360-365

THE APPLICATION OF THE THEORY OF PLANNED BEHAVIOUR TO CONSUMER FOOD CHOICE

R. Shepherd, Institute of Food Research

P. Sparks, Institute of Food Research

C.A. Guthrie, Institute of Food Research

ABSTRACT -

Human food choice is a complex phenomenon influenced by a wide range of factors. The theory of planned behaviour offers a means for trying to understand the roles of some of these factors. Evidence will be presented for the inclusion of measures of self-identity and moral obligation. Problems with the measurement of perceived control and possible means for overcoming them will also be discussed. Future research will include the application of the causal modelling technique LISREL and collaborative work with Aarhus and Helsinki seeking to establish the cross-cultural validity of such approaches.

MODELS OF FOOD CHOICE

The choice of foods is an area of concern for many people involved in the production and distribution of foods, and for those concerned with nutrition and health education. Relatively little is known about how and why people choose the foods that constitute their diets or about how their choices can be influenced in an effective way.

Like any complex human behaviour, food choice will be influenced by many factors. There are a number of models in the literature, which seek to delineate the effects of likely influences (e.g. Pilgrim, 1957; Khan, 1981; Randall and Sanjur, 1981; Shepherd, 1985; reviewed by Shepherd, 1989). However, many such models are simply catalogues of the likely influences. Few of them present any indication of the likely mechanisms of action of the multitude of factors identified, nor do they quantify the relative importance of factors or allow any quantitative tests which are predictive of food choice. Although such models can be useful in pointing to the variables to consider in studies in this area, they do not provide a framework for quantitative modelling of food choice behaviour.

An example of such a model is shown in Figure 1. The factors influencing food choice are categorised as those related to the food, to the person making the choice and to the external economic and social environment within which the choice is made. Some chemical and physical properties of the food will be perceived by the person in terms of sensory attributes, e.g. flavour, texture or appearance. However, perceiving these sensory attributes in a particular food does not necessarily mean that a person will or will not choose to consume that food. It is the person's liking for that attribute in that particular food which will determine whether or not the food is chosen. Other components in the foods will have effects upon the person, e.g. reducing hunger, and the learning of the association between the sensory attributes of a food and its post-ingestional consequences appears to be a major mechanism by which preferences develop. Marketing and economic variables, as well as social, cultural, religious or demographic factors are also likely to be very important (Murcott, 1989; Shepherd, 1989).

THE THEORY OF PLANNED BEHAVIOUR

Many of the influences on food choice are likely to be mediated by the beliefs and attitudes held by an individual. Thus beliefs about the nutritional quality and health effects of a food may be more important than actual nutritional quality and health consequences in determining a person's choice. Likewise various marketing, economic, social, cultural, religious or demographic factors will act through the attitudes and beliefs held by the person. As such, the study of the relationship between choice and the beliefs and attitudes held by a person offers one possible route towards a better understanding of the influence of different factors on food choice.

Attitudes are assumed to be causally related to behaviour. This is true both in the common use of the term attitude and within the research literature in social psychology (Eagly and Chaiken, 1993), but the empirical evidence for this link has not always been clear. For example, in the nutrition literature many studies have attempted to measure the association between attitudes and consumption of foods. Axelson, Federline and Brinberg (1985) performed a meta-analysis of such studies and found evidence for a small (although statistically significant) correlation between attitudes and behaviour (r=0.18). Thus a superficial survey of this area might lead to the conclusion that attitudes are not related to behaviour to any important degree. The same type of finding in social psychology led to a crisis in attitude research in the late 1960s (Wicker, 1969), which resulted in the generation of a number of structured attitude models (e.g. Ajzen and Fishbein, 1980). One example of such a structured attitude model is the theory of reasoned action (TRA) (Ajzen and Fishbein, 1980) and its extension in the form of the theory of planned behaviour (TPB) (Ajzen, 1988, 1991). These models have been applied to a wide range of behaviours, including consumer behaviour, and more recently have been applied to food choice issues.

The TRA seeks to explain rational behaviour over which people have complete control. The TPB additionally seeks to explain non-volitional behaviours, goals and outcomes, which are not entirely under the control of the person. With volitional behaviours it is argued that intention to perform a behaviour is the best predictor of behaviour. Intention, in turn, is predicted by two components: the person's own attitude (e.g. whether the person sees the behaviour as good, beneficial, pleasant, etc.) and perceived social pressure to behave in this way (termed the subjective norm). The TPB includes a component of perceived control, which is hypothesised, along with attitude and subjective norm, to predict behavioural intentions and may also influence the intention-behaviour link (see Figure 2).

Attitude, in turn, is predicted by the sum of products of beliefs about outcomes of the behaviour and the person's evaluations of these outcomes as good or bad. The subjective norm is predicted by the sum of products of normative beliefs, which are perceived pressure from specific influential people or groups (e.g. doctors, family) and the person's motivation to comply with the wishes of these people or groups. In a similar fashion, perceived control is determined by the sum of specific control beliefs modified by the perceived power of the control factors to facilitate or inhibit performance of the behaviour (Ajzen, 1991).

The original conception of the TRA assumed that influences other than beliefs, attitudes, social pressure and intention would act through these variables (Ajzen and Fishbein, 1980); this would also apply to the TPB, although here perceived control would be an additional component of the model. Thus demographic variables, such as age or social class, should influence behaviour only through the model variables and not act as independent influences on behaviour.

FIGURE 1

SOME FACTORS AFFECTING FOOD CHOICE AND INTAKE (FROM SHEPHERD, 1985)

The TRA has been widely applied to many issues in social psychology (Ajzen and Fishbein, 1980; Tesser and Shaffer, 1990), to issues such as blood donation (Bagozzi, 1981), coupon usage (Shimp and Kavas, 1984), contraception use (Davidson and Morrison, 1983) and seat belt use (Wittenbraker, Gibbs and Kahle, 1983).

The TRA has also been successfully applied to a range of food choice issues, including eating in fast food restaurants (Axelson, Brinberg and Durand, 1983), beef consumption (Sapp and Harrod, 1989), salt intake (Shepherd and Farleigh, 1986), choice of milk of different fat levels (Tuorila, 1987; Shepherd, 1988), choice of high fat foods (Tuorila and Pangborn, 1988), fat intake (Shepherd and Stockley, 1985, 1987; Shepherd and Towler, 1992; Towler and Shepherd, 1992) and adolescent food choice (Dennison and Shepherd, 1995).

Sheppard, Hartwick and Warshaw (1988) carried out a meta-analysis of 87 studies using this model in the area of general consumer choice (not specifically related to foods). They found an estimated correlation of 0.53 between intention and behaviour and a multiple correlation of 0.66 between attitude plus subjective norm against intention (Sheppard, Hartwick and Warshaw, 1988). This therefore suggests that the model has validity both in the study of general consumer choice and the study specifically of food choice.

The inclusion of perceived control has received some empirical support, for example in predicting the intention of mothers to limit the sugar consumption of children (Beale and Manstead, 1991), and in studies of weight loss (Schifter and Ajzen, 1985) and dietary health behaviours (Ajzen and Timko, 1986). It has, however, not been found to improve prediction of intentions or behaviour in all applications (Fishbein and Stasson, 1990). In a study of biscuit and bread consumption (Sparks, Hedderley and Shepherd, 1992), intentions to consume wholemeal bread were not influenced by perceived control but intentions to consume biscuits were. Thus, inclusion of a measure of perceived control may be important in predicting choices of some, although not all, foods.

Although the TRA and TPB have been applied successfully in the food choice area there are a number of shortcomings in their conceptualisation and implementation. This has led to a number of suggested modifications and extensions. Two such extensions will be described here, as applied to issues of food choice.

THE ROLE OF SELF-IDENTITY IN THE TPB

One recent suggested modification to the TPB is that a person's self-identity may influence behaviour independently of his or her attitudes (Biddle, Bank and Slavinge, 1987; Charng, Piliavin and Callero, 1988). In the field of blood donation, a person's identification of him or herself as a blood donor has been found to be a predictor of intention to donate blood over and above the effects of the person's attitude towards blood donation (Charng, Piliavin and Callero, 1988). While it would be expected that people who see or do not see themselves as blood donors would differ in their beliefs and attitudes concerning blood donation, it would seem unlikely that such a self image would impact on intention independent of attitudes. Rather the effects of self-identity would be expected to act through differences in the beliefs and attitude component of the TPB.

The inclusion of self-identity in the TPB was investigated in a study of the consumption of organic vegetables (Sparks and Shepherd, 1992). Two hundred and sixty-one subjects completed questionnaires which included measures of the components of the TPB, along with two measures of identification with green consumerism: 'I think of myself as a green consumer' and 'I think of myself as someone who is very concerned with green issues' (with responses from 'disagree very strongly' to 'agree very strongly'). The responses from these two measures were summed to create a measure of self-identity.

Correlations between the components of the TPB were reasonably high, confirming the basic applicability of the model. The summed products of beliefs and evaluations correlated significantly with attitudes (r=0.44, p<0.001). A multiple regression tested the hypothesis that self-identity would not add to the prediction of intentions over and above the contributions of the other components. The multiple regression (R=0.52, p<0.001) had significant beta coefficients for attitude (beta=0.21, p<0.01), subjective norm (beta=0.16, p<0.05) and perceived control (beta=0.26, p<0.001). However, self-identity also significantly added to this regression (beta=0.22, p<0.01).

FIGURE 2

SCHEMATIC REPRESENTATION OF THE THEORY OF PLANNED BEHAVIOUR

One possible reason for the self-identity measure having such an independent effect is that the people may have been using responses to this question as a proxy for past behaviour, i.e. inferring their self-identity response from an examination of past behaviour, and past behaviour has been found to be a powerful predictor of both intention and future behaviour (Bentler and Speckart, 1979). However, even after adding responses on past behaviour into the above regression the measure of self-identity was still found to be an independent predictor of intention, thus arguing against this interpretation of the effect.

The reason for the independent effect of self-identity is not entirely clear (Sparks and Shepherd, 1992), although it may reflect the inability of current attitude measures adequately to assess various symbolic and emotional factors which may nonetheless influence intention and behaviour. Extensions of this kind to the basic model offer a potential means towards a better understanding of the different factors determining food choice.

EXTENSION OF THE TPB TO INCLUDE MORAL OBLIGATION

Another aspect of behaviour inadequately assessed by existing components of the TPB is that of moral or ethical concerns. In its basic form the model is purely utilitarian or instrumentally based with behaviour leading to outcomes which are viewed as either positive or negative. For some behaviours, moral considerations might also have a significant impact; some of these considerations may be independent of instrumental beliefs about outcomes. Perceived moral obligation has been investigated in studies outside the food area, primarily on behaviours such as stealing or cheating where a moral component might be expected to operate (e.g. Beck and Ajzen, 1991). In such cases it has been found to add significantly to the predictive power of the basic TPB.

Although food choice is a less obvious domain within which such factors might operate there are nonetheless particular instances where moral or ethical issues might be important. These might include the use of animals in food production and issues of animal welfare or the application of new techniques to food production, such as genetic engineering. Another possible area of moral concern involves situations where decisions are made on behalf on others and the person making the decision might then feel a moral obligation toward protecting those affected by the decision. Thus within families, a perceived responsibility rests with the person making the choice of foods for the health and well-being of other family members and this may be thought of as a perceived moral obligation for promoting the health of family members.

This issue has been examined in a study of the consumption of different types of milks (Raats, Shepherd and Sparks, 1995). Two hundred and fifty seven people completed a questionnaire on the consumption of whole, semi-skimmed and skimmed milk. In addition to questions assessing the traditional components of the TPB there was one moral obligation measure for each type of milk: 'I feel obliged to use skimmed/semi-skimmed/whole milk for my family's health', with responses on 9-category scales ranging from 'disagree very strongly' to 'agree very strongly'.

For each type of milk the basic model was shown to give a reasonable degree of prediction of intention. Perceived control added significantly to the prediction of intention to consume skimmed milk (p<0.001) but was only marginally significant in the prediction of intention to consume semi-skimmed milk (p=0.06) and not significant for whole milk. This confirms the applicability of the perceived control component, with predictions over and above the impact of attitude and subjective norm for some, but not all, choices.

TABLE 1

RESULTS FROM MULTIPLE REGRESSIONS PREDICTING ATTITUDE TO CONSUME MILK OF DIFFERENT FAT LEVELS. THE MULTIPLE CORRELATION COEFFICIENTS (R) ARE SHOWN FOR ENTERING FIRST THE BELIEF-EVALUATIONS AND THEN THE MEASURES OF MORAL OBLIGATION; THE BETA COEFFICIENTS ARE FROM THE FINAL REGRESSION EQUATION.

When the responses on obligation for family's health were added into the regression predicting intention, this led to small but significant increases in prediction for whole and semi-skimmed milk (p<0.05) but not for skimmed milk. If, however, we consider the moral obligation component as acting a stage further back in the model and feeding into attitude in addition to beliefs and evaluations, we get the results shown in Table 1. As expected, belief-evaluations are highly significant predictors of attitude in each case but the measure of obligation for family's health has an effect over and above this and is highly significant for each type of milk.

Thus moral obligation appears to be of some importance within the TPB but it is suggested that its position in the model (for these behaviours) may not lie at the level of being an extra predictor of intention but rather may be thought of as a stage further back, feeding into attitude in addition to belief-evaluations about the outcomes of the behaviour. Such moral obligation questions appear to tap something which is not measured by questions on beliefs concerning instrumental outcomes of actions. Although previously shown to be true for behaviours with an obvious moral component, such as stealing or cheating, the demonstration of its importance in the choice of foods suggests it may have wider applicability. Like perceived control and self-identity, it is unlikely to be important for the choice of all types of foods in all contexts but it may be particularly important when food choices are made on behalf of others, and perhaps particularly for children.

THE MEASUREMENT AND NATURE OF PERCEIVED CONTROL

As mentioned above, there has been much empirical evidence to support the inclusion of perceived control as part of the basic model structure. However, problems remain regarding the conceptualisation and the operationalisation of this component. Perceived control is typically measured by summing items (normally 3 or 4 questions) which ask people to make judgements regarding the amount of control they have over a behaviour or the amount of difficulty they experience in carrying out the behaviour. A number of authors (Beale and Manstead, 1991; Dennison and Shepherd, 1995) have commented on the poor inter-item reliability between such items and have highlighted the conceptual differences between the issue of control and the issue of difficulty.

These issues have recently been addressed in studies examining people's consumption of red meat and chips. Using perceived control questions taken from the literature, TPB questionnaires were administered to two groups of subjects (Study 1, n=91; Study 2, n=97). Principal components analysis of the perceived control items revealed a distinct split between the different kinds of items, with items measuring 'difficulty/ease' issues loading on one factor, and items measuring 'control' issues loading on a separate factor. The analysis also found that measures of 'difficulty/ease' had a greater predictive effect on intention than measures of 'control'. These findings support the argument for a redefinition of the perceived control component, allowing improved methods of measurement to be developed.

CURRENT RESEARCH ON APPLYING THE TPB TO FOOD CHOICE

The TPB and extensions of its basic form have been successfully applied to a number of food choice problems but there are still outstanding issues to be considered. Many of these issues will be explored as part of a 3 year EC-funded project on "The development of models for understanding and predicting consumer food choice". The research on the TPB will be carried out primarily at the Institute of Food Research in the UK but will also involve collaboration with the Business School at Aarhus, Denmark and the University of Helsinki, Finland.

The aims of this research project are to test the applicability of the self-identity concept, to examine the structure of the perceived control component and to examine the beliefs which act as pre-cursors of perceived control. Alternative data analysis procedures will be developed to test these relationships, in particular the application of LISREL. Finally the validity of the models developed will be tested across the three European countries.

Data gathered using the TPB are usually analysed by means of correlation and multiple regression. It has been suggested by several authors that the LISREL model may provide a better approach (Bagozzi and Yi, 1989). LISREL involves modelling of the matrix of covariance between the items of the questionnaire. It is similar to factor analysis, in that it postulates that the responses are indicators of unobservable variables (latent variables or factors). LISREL allows both the construction of a model incorporating these latent variables and the testing of the model through the covariances between their indicators. In a sense LISREL is also close to regression, but it does not require that the independent variables are recorded without random errors as is assumed under regression. Using LISREL, the measurement errors can be separated from the theoretical constructs with the use of multiple measures. In addition, the interactions among constructs can be tested simultaneously, and different mediating factors can be included.

LISREL has been applied successfully in marketing and in economics, but has yet to be fully exploited in studies of human food choice. It offers an excellent method for integrating the effects of different types of variables and provides a quantitative method for determining the appropriateness of models tested. It will be applied to data generated in the current project in order to test the importance of variables in influencing food choice.

Ideally, research on behaviour associated with food aims at findings of cross-cultural validity but most studies of food choice are specific to one particular country and culture. Although it can be hoped that the models and approaches developed are potentially more widely applicable and generalisable, the cross-cultural validity of social science measurements cannot usually be taken for granted. This is in part due to semantic differences in the languages used, even after careful translations. The aim within this project is to test the validity of the extended TPB across states in predicting the choice of foods. By providing parallel measurements in various countries an attempt will be made to establish the cross-cultural validity of the models developed based on the TPB.

Each of participants will be involved in a joint multi-centre study testing the applicability of the TPB across countries and languages on the same range of foods. A standardised battery of methods for the measurement of the most crucial factors influencing food choice will be developed and these will be translated for application in each country. This will involve back translation and checking in order to ensure comparability across languages. These methods will be utilised in studies using the same protocols in three different countries with different languages in order to compare factors influencing food choice.

CONCLUSIONS

Food choice is determined by a large range of potential factors. Many of the models put forward in this area simply list possible influences rather than offering frameworks for empirical research and practical application. Although there is general agreement on the types of influences likely to be important, the integration of these factors into a coherent and quantitative model of food choice remains an area in need of development.

The TPB offers a starting point, but the component of perceived control requires a clearer formulation and operationalisation. There are various extensions of the basic model, including self-identity and moral obligation, which potentially offer a means for developing a clearer understanding of the factors influencing the choice of specific types of foods in specific contexts. In particular they make possible the exploration of the more emotional and affective elements potentially important in food choice, rather than simply addressing the rational, cognitive issues prevalent in the literature.

ACKNOWLEDGEMENTS

The work reported here was funded by the Biotechnology and Biological Sciences Research Council and the European Community.

REFERENCES

Ajzen, I. (1988), Attitudes, Personality, and Behavior, Milton Keynes: Open University Press.

Ajzen, I. and M. Fishbein (1980), Understanding Attitudes and Predicting Social Behavior, Englewood Cliffs, New Jersey: Prentice-Hall.

Ajzen, I. and C. Timko (1986), "Correspondence Between Health Attitudes and Behavior," Basic and Applied Social Psychology, 7, 259-276.

Axelson, M.L., D. Brinberg and J.H. Durand (1983), "Eating at a Fast-food Restaurant - A Social-psychological Analysis," Journal of Nutrition Education, 15, 94-98.

Axelson, M.L., T.L. Federline and D. Brinberg (1985), "A Meta-analysis of Food and Nutrition-related Research," Journal of Nutrition Education, 17, 51-54.

Bagozzi, R.P. (1981), "Attitudes, Intentions, and Behavior: A Test of Some Key Hypotheses," Journal of Personality and Social Psychology, 41, 607-627.

Bagozzi, R.P. and Y. Yi (1989), "On The Use of Structural Equation Models In Experimental Designs," Journal of Marketing Research, 26, 271-284.

Beale, D.A. and A.S.R. Manstead (1991), "Predicting Mothers' Intentions to Limit Frequency of Infants' Sugar Intake: Testing the Theory of Planned Behavior," Journal Applied Social Psychology, 21, 409-431.

Beck, L. and I. Ajzen (1991), "Predicting Dishonest Actions Using the Theory of Planned Behavior," Journal of Research in Personality, 25, 285-301.

Bentler, P.M. and G. Speckart (1979), "Models of Attitude-behavior Relations," Psychological Review, 86, 452-464.

Biddle, B.J., B.J. Bank and R.L. Slavinge (1987), "Norms, Preferences, Identities and Retention Decisions," Social Psychology Quarterly, 50, 322-337.

Charng, H.-W., J.A. Piliavin and P.L. Callero (1988), "Role Identity and Reasoned Action in the Prediction of Repeated Behavior," Social Psychology Quarterly, 51, 303-317.

Davidson, A.R. and D.M. Morrison (1983), "Predicting contraceptive behavior from attitudes: A comparison of within- versus across-subjects procedures," Journal of Personality and Social Psychology, 45, 997-1009.

Dennison, C.M. and R. Shepherd (1995), Adolescent Food Choice: An Application of the Theory of Planned Behaviour," Journal of Human Nutrition and Dietetics, 8, 9-23.

Eagly, A.H. and S. Chaiken (1993), The Psychology of Attitudes, San Diego: Harcourt, Brace and Jovanovich.

Fishbein, M. and M. Stasson (1990), "The Role of Desires, Self-predictions, and Perceived Control in the Prediction of Training Session Attendance," Journal of Applied Social Psychology, 20, 173-198.

Khan, M.A. (1981), "Evaluation of Food Selection Patterns and Preferences," CRC Critical Reviews in Food Science and Nutrition, 15, 129-153.

Murcott, A. (1989), "Sociological and Social Anthropological Approaches to Food and Eating," in World Review of Nutrition and Dietetics 55, ed. G.H. Bourne, Basel: Karger, pp. 1-40.

Pilgrim, F.J. (1957), "The Components of Food Acceptance and their Measurement," American Journal of Clinical Nutrition, 5, 171-175.

Raats, M.M., R. Shepherd and P. Sparks (1995), "Including Moral Dimensions of Choice Within the Structure of the Theory of Planned Behavior," Journal of Applied Social Psychology, 25, 484-494

Randall, E. and D. Sanjur (1981), "Food Preferences - Their Conceptualization and Relationship to Consumption," Ecology of Food and Nutrition, 11, 151-161.

Sapp, S.G. and W.J. Harrod (1989), "Social Acceptability and Intentions to Eat Beef: An Expansion of the Fishbein-Ajzen Model using Reference Group Theory," Rural Sociology, 54, 420-438.

Schifter, D.E. and I. Ajzen (1985), "Intention, Perceived Control, and Weight Loss: An Application of the Theory of Planned Behavior," Journal of Personality and Social Psychology, 49, 843-851.

Shepherd, R. (1985), "Dietary Salt Intake," Nutrition and Food Science, 96, 10-11.

Shepherd, R. (1988), "Belief Structure in Relation to Low-fat Milk Consumption," Journal of Human Nutrition and Dietetics, 1, 421-428.

Shepherd, R (1989), "Factors Influencing Food Preferences and Choice," in Handbook of the Psychophysiology of Human Eating, ed. R. Shepherd, Chichester: Wiley, pp. 3-24.

Shepherd, R. and C.A. Farleigh (1986), "Preferences, Attitudes and Personality as Determinants of Salt Intake," Human Nutrition: Applied Nutrition, 40A, 195-208.

Shepherd, R. and L. Stockley (1985), "Fat Consumption and Attitudes Towards Food with a High Fat Content," Human Nutrition: Applied Nutrition, 39A, 431-442.

Shepherd, R. and L. Stockley (1987), "Nutrition Knowledge, Attitudes, and Fat Consumption," Journal of the American Dietetic Association, 87, 615-619.

Shepherd, R. and G. Towler (1992), "Nutrition Knowledge, Attitudes and Fat Intake: Application of the Theory of Reasoned Action," Journal of Human Nutrition and Dietetics, 5, 387-397.

Sheppard, B.H., J. Hartwick and P.R. Warshaw (1988), "The Theory of Reasoned Action: A Meta-analysis of Past Research with Recommendations for Modifications and Future Research," Journal of Consumer Research, 15, 325-343.

Shimp T.A. and A. Kavas (1984), "The Theory of Reasoned Action Applied to Coupon Usage," Journal of Consumer Research, 11, 795-809.

Sparks, P., D. Hedderley and R. Shepherd (1992), "An Investigation into the Relationship between Perceived Control, Attitude Variability and the Consumption of Two Common Foods," European Journal of Social Psychology, 22, 55-71.

Sparks, P. and R. Shepherd (1992), "Self-identity and the Theory of Planned Behavior: Assessing the Role of Identification with "Green Consumerism," Social Psychology Quarterly, 55, 388-399.

Tesser, A. and D.R. Shaffer (1990), "Attitudes and Attitude Change," Annual Review of Psychology, 41, 479-523.

Towler, G. and R. Shepherd (1992), "Application of Fishbein and Ajzen's Expectancy-value Model to Understanding Fat Intake," Appetite, 18, 15-27.

Tuorila, H. (1987), "Selection of Milks with Varying Fat Contents and Related Overall Liking, Attitudes, Norms and Intentions," Appetite, 8, 1-14.

Tuorila, H. and R.M. Pangborn (1988), "Behavioural Models in the Prediction of Consumption of Selected Sweet, Salty and Fatty Foods," in Food Acceptability, ed. D.M.H. Thomson, London: Elsevier Applied Science, pp. 267-279.

Wicker, A.W. (1969), "Attitude Versus Actions: The Relationship of Verbal and Overt Behavioral Responses to Attitude Objects," Journal of Social Issues, 25, 41-78.

Wittenbraker, J., B.L. Gibbs and L.R. Kahle (1983), "Seat Belt Attitudes, Habits, and Behaviors: An Adaptive Amendment to the Fishbein Model," Journal of Applied Social Psychology, 13, 406-421.

----------------------------------------