Testing Competing Models of Consumer Decision Making in the Preventive Health Care Marketplace

ABSTRACT - Three models borrowed from social psychology, medical sociology, and communication theory which have been proposed to predict behavioral intention were compared on the basis of their conceptual strength and predictive accuracy in a study of a community swine flu vaccination program. The results suggest that models incorporating evaluative components, normative influences, and intervening summary concepts may yield a greater understanding of health care decisions. Implications of the findings were discussed.


Richard L. Oliver and Philip K. Berger (1978) ,"Testing Competing Models of Consumer Decision Making in the Preventive Health Care Marketplace", in NA - Advances in Consumer Research Volume 05, eds. Kent Hunt, Ann Abor, MI : Association for Consumer Research, Pages: 277-282.

Advances in Consumer Research Volume 5, 1978      Pages 277-282


Richard L. Oliver, University of Iowa

Philip K. Berger, University of Kentucky

[This research was partially supported by grants from the University of Kentucky Graduate School and the Lexington-Fayette County Health Department.]


Three models borrowed from social psychology, medical sociology, and communication theory which have been proposed to predict behavioral intention were compared on the basis of their conceptual strength and predictive accuracy in a study of a community swine flu vaccination program. The results suggest that models incorporating evaluative components, normative influences, and intervening summary concepts may yield a greater understanding of health care decisions. Implications of the findings were discussed.


The application of behavioral concepts to health product and service marketing was a natural outgrowth of the broadened marketing concept (Kotler and Levy, 1969). Early attempts to construct comprehensive models (Zaltman and Vertinsky, 1971; Becker, 1974; Rosenstock, 1974) were followed by various applications in medical sociology (Cf. Becker and Maiman, 1975, for review) and a recent symposium on behavior in the health marketplace (Newman, 1976). Unfortunately, consumer behaviorists have been slow to apply the emerging knowledge gleaned from studies on non-health related products to the medical field.

The purpose of this paper is to compare the extent to which a model currently used by medical sociologists to explain preventative health behavior and a well accepted behavioral intention model used in the marketing community can predict intentions to receive the swine flu vaccination. In addition, a recent hybrid model focusing on drive arousal and reduction tendencies will also be contrasted with the other two models. Hopefully, these comparisons will permit medical care and public health researchers to modify their present perceptions of preventative health and health compliance behavior to include conceptual modifications suggested in the social-psychological and consumer behavior areas. Moreover, it will also be shown that many of the concepts used in health behavior models are isomorphic with the more traditional views of market behavior.

The Models

The three models to be tested include Fishbein's (Fishbein and Ajzen, 1975) Behavioral Intention Model (BIM), the Health Belief Model (HBM) as discussed by Becker and Maiman (1975), and a drive component model proposed by Wortzel (1976) to predict health care decisions. All are to some degree "instrumentality" schemata as they suggest that action is undertaken because it is perceived to result in or block various life outcomes. Each model is specified in some detail in the following sections.

The Behavioral Intention Model. Because consumer behaviorists will be most familiar with Fishbein's BIM, only a brief discussion is presented here. Based on Dulany's (1968) theory of propositional control, Fishbein (1967; Fishbein and Ajzen, 1975) proposed that one's intention to perform a specific behavioral action is an additive function of two components, attitude toward the behavior and normative social influences. One's attitude toward the behavior, in turn, is a function of the sum of products of beliefs about the consequences of the behavior and one's evaluation of those consequences. In similar fashion, the summary perception of normative social influences, referred to as one's subjective norm, is believed to be a function of the sum of products of normative beliefs attributed to various referent persons and one's motivation to comply with those beliefs. Other stimulus conditions are thought to influence intention through the attitudinal and normative components.

The Health Belief Model. As summarized by Becker and Maiman (1975), the HBM predicts that people will engage in preventive health behavior if "they possess minimal levels of relevant health motivation and knowledge, perceive themselves as potentially vulnerable and the condition threatening, are convinced of the efficacy of intervention, and see few difficulties in undertaking recommended action (p. 12)." Unlike Fishbein's Behavioral Intention Model, the HBM is less formalized in its constituent components and the measurement of its variables. Based primarily upon articles by Becker (1974), Becker and Maiman (1975), Maiman and Becker (1974), and Rosenstock (1974), the version of the HBM discussed in the next section has been adopted for the purpose of this study.

Proponents of the HBM suggest that the probability of an individual engaging in a preventive health action is a function of three factors, a "benefits-barriers" analysis of the advantages and disadvantages of the health prevention activity, the "perceived threat" associated with the condition or illness, and various "cues to action" which include both mass media and interpersonal communications. [The model described here departs in one important respect from earlier versions (e.g., Rosenstock, 1974). The cues to action component in prior formulations impacted directly on the perceived threat component and indirectly on the likelihood of behavior variable. A careful reading of the literature however, suggests that cues to action more appropriately affect the likelihood of behavior directly rather than through perceived threat. This latter interpretation is the one tested in this paper.] The perceived threat construct, in turn, is believed to have two antecedents, namely one's perceived susceptibility to the health problem and the anticipated severity of the problem if contracted. In a manner similar to that employed in the BIM model, other determinants are thought to affect the antecedents of one's intention rather than the intention itself.

In summary, an individual's perceptions of susceptibility to and severity of the focal health problem constitute the inherent threat posed by the disease. Given that a preventive health action is available, the model assumes that the person evaluates the action in terms of its perceived potential benefits to reduce this threat as well as the potential barriers or costs to taking the same action. Finally, the individual is also assumed to be further influenced by personal and impersonal recommendations. These three major components are believed to define one's subjective state of readiness to take preventive action. [Some versions of the model suggest that demographic and personality variables affect perceptions of the major variables. These individual difference constructs have been purposely omitted in order to focus on the cognitive concepts in the HBM.]

A Comparison of the BIM and HBM

A number of differences exist between Fishbein's BIM and the HBM. Some have been suggested by Jaccard (1975), others derive from exchanges between Songer-Nocks (1976) and Fishbein and Ajzen (1976) while some interpretations are novel to this paper. Each is summarized briefly in the following discussion.

First, Fishbein's BIM is designed to predict intention to act, not action itself. The original version of the HBM contains no variable similar to behavioral intention but is designed to predict actual behavior. If this difference is taken at face value, it would not be possible to compare the two models. For the purposes of this paper, however, the criterion for both models was identified as behavioral intention. If one is willing to assume a high positive correlation between intention and behavior, the data are useful for comparing the two models. [The relationship between intention and behavior is an empirical issue which the authors intend to address at a future time when behavioral data collected by the local health department can be retrieved.]

Second, the BIM requires that beliefs concerning the consequences of the preventive health act be multiplicatively combined with one's evaluation of those consequences. In the HBM, evaluations of the consequences of the act are not explicitly called for and, as a result, may not be measured. Since the HBM is not as formally specified as the BIM, the degree of precision surrounding the explication of consequences is not known. It can be said, however, that the outcomes of taking action are assessed as perceived benefits and barriers. Moreover, it appears that all benefits are assumed to be universally evaluated as good while the barriers are assumed to be consistently evaluated as bad.

Third, social norms are a necessary component of Fishbein's model. In contrast, the analogous normative influences in the HBM are listed among many other variables in the cues to action construct. By the same line of reasoning, the BIM conceptualization does not explicitly provide for various mass media messages but rather subsumes them under the category of exogenous stimulus conditions. Thus, depending on the manner in which a particular investigator operationalizes influences external to the individual in either model, the relevant social norms or media impacts may not be included.

A fourth difference between the two models concerns the fact that the HBM maintains, as separate components, the perceived susceptibility, severity, and threat of the potential illness. Because Fishbein's model was developed in non-health related environments, the need to explicitly provide for emotional fear arousal variables was not evident. Consequently, the BIM is limited, to a certain extent, to the rational side of a preventive health decision. Although the HBM does not explicitly state the rules for combining susceptibility and severity to arrive at perceived threat, it does provide for their specification in the model.

To aid the reader in comparing the two instrumentality models described in the preceding discussion, the constructs in the BIM and HBM are juxtaposed on the basis of their isomorphic similarities in Table 1 along with the components of the Drive Reduction Model (DRM). Discussion of this latter model, due to its somewhat different orientation, has been reserved until this time.



The Drive Reduction Model. A preventive health model based on the fear arousing properties of an illness and the capacity of a health action to reduce that fear has been proposed by Wortzel (1976) as adapted from earlier work by Bauer and Cox (1963) on the rational and emotional content of communication influences. Bauer and Cox originally suggested that the two critical dimensions which affect a consumer's motivation to engage in a given act are the level of emotional arousal resulting from failure to take action and the subjective probability that that same action will prevent the anticipated problem which is generating this arousal. The consumer's level of felt emotion is considered to be the drive arousal dimension while the subjective degree of success attached to a preventive action is believed to be the drive reduction mechanism.

The authors predict that, when the level of the subjective probability of success relative to arousal is sufficiently high, a person will be motivated to engage in the act; when that contrast goes below a certain (unstated) value, the person will choose not to engage in that behavior.

Wortzel (1976) later operationally redefined the Bauer and Cox (1963) drive arousal and drive reduction dimensions in term of the variables encompassed by the HBM. Specifically, the drive arousal dimension was conceptualized as the product of an individual's perception of his susceptibility to a disease and the perceived seriousness of the same disease. The drive reduction dimension, in turn, was defined as the difference between the subjective probability that a preventive health action will result in avoiding the disease and one's subjective assessment of the cost of preventive behavior. Predictions from this reoperationalized model are identical to those of Bauer and Cox, i.e., action is a function of the level of drive reduction as related to drive arousal.

Table 1 shows how the variables suggested in the DRM compare to those discussed with regard to the BIM and HBM. Further elaboration is necessary, however, because a number of subtle differences exist in the variables used to construct the major concepts in the model. First, the number of outcomes of the preventive behavior used in the DRM is somewhat restricted. Only one positive consequence, that of immunity from the disease, is specified while the negative outcomes are limited to "costs" which are subjectively identified. Thus, the DRM may exclude a number of positive outcomes (e.g., protecting one's family from contagion) or negative outcomes such as uncomfortable, unknown, side effects. In addition, the subject is not asked to make evaluations of the degree of immunity received or of the costs, a problem which is also inherent in the HBM.

Second, the DRM is more limited and also more explicit in its operational definition of perceived susceptibility to and perceived seriousness of the disease. While the HBM admits many diverse operational definitions of these variables (Becker and Maiman, 1975), the DRM requires a probabilistic judgment of susceptibility in the absence of the preventive health behavior. In an equally rigorous manner, seriousness is viewed as a combination of all ill effects of the disease. Also of note is the fact that the DRM specifies that the combinatorial rule for these two variables is multiplicative. Jaccard (1975) has noted that in various formulations of the HBM, both additive and multiplicative rules have been used.

Finally, the DRM appears to suffer from a greater degree of conceptual deficiency than either of the other two models. In addition to the previously discussed omissions, no measures of personal or mass media influences are specified nor is the summary construct of perceived threat. Rather, these variables appear to have been considered either redundant, exogenous, or uncontrollable.


The degree to which the three models predict one's intention to receive a swine flu vaccination is the central focus of this paper. The corresponding null hypothesis specifies that no differences will be observed. The next section discusses the methodology employed to test this notion.



A cross-sectional field study of a community swine flu vaccination program was undertaken in October 1976. Questionnaires measuring beliefs, attitudes, and intentions concerning the swine flu inoculation campaign were mailed to residents one week before the vaccine first became available in that community. A number of other municipalities across the country had already initiated their vaccination programs.


Two thousand residents of a medium-size Midwestern community were selected from the telephone directory using a systematic random sampling procedure to receive the questionnaire used in the study. In addition, 1000 students from a major state university located in the community were also selected by the same random process from the student directory to participate in the survey.


Behavioral Intention Model. Fishbein (1975) requires that all salient consequences of the behavioral referent as well as the relevant reference persons be identified. This task was facilitated by the controversial nature of the swine flu inoculation campaign. Because of the widespread media coverage of the swine flu program, the pros and cons of immunization were fairly well documented (e.g., Boffey, 1976; Spivak, 1976). In addition, a culling of varied media presentations including local TV interviewing provided a sufficient list of consequences and referent persons and groups. The number of factors in each list was later reduced through pretesting to eight consequences of obtaining the swine flu shot, five consequences of failing to obtain the shot, and five social referents. [The consequences of obtaining the shot were a mild case of the flu, pain from the shot, immunity from the flu, protecting family and friends from the flu, being inconvenienced, side effects or allergenic reactions, feeling more comfortable with people, and possible fatality due to the shot. The consequences of not getting the shot included catching swine flu and, if one caught the flu, pain and discomfort, cost of medicine and doctor's bills, lost time from one's job and/or school, and death. The social referents were one's doctor, family or spouse, local health department, company or boss, and friends and neighbors.]

To obtain a probabilistic measure of beliefs about the consequences, the respondents were asked to scale the possibility of occurrence of each consequence if they got (did not get) a swine flu shot on a five point scale ranging from "no chance" (0) through "50-50" (.5) to "certain" (1). The evaluative component of the model was measured by asking respondents to evaluate each consequence on a five point good-bad scale ranging from "very bad" (-2) to "very good" (+2). In a similar manner, subjects indicated their normative beliefs with regard to each social referent by responding to the question "Do the following people want you to get a swine flu shot?", on a five point scale ranging from "yes" (+2) to "no" (-2). Finally, motivation to comply was obtained from four point items asking the respondent: "Does (social referent's) opinion matter to you?" Scale points ranged from "yes" (3) to "no" (0).

Summary constructs tapping one's overall attitude toward getting a swine flu shot and one's overall subjective norm were measured on independent scales. The attitude toward the act measure used in the study was a nine item semantic differential scale [The coefficient alpha reliability of this scale was .94.] while the subjective norm variable was a one item summary measure suggested by Fishbein and Ajzen (1975, p. 314).

The Health Belief Model. Perceived susceptibility, severity, and threat were measured on five, eight, and four item Likert scales respectively. Items for these scales were suggested by Rosenstock's (1974) review of earlier studies and by media statements specific to the swine flu campaign. [The scales yielded coefficient alpha reliabilities of .68, .72, and .70 respectively.]

To obtain a measure of the benefits-barriers concept, scores from the probability of consequences scale in the BIM (i.e., the belief component) were used without the evaluative dimension. Thus, the average of the probabilities assigned to the five unfavorable outcomes (barriers) was subtracted from the mean of probabilities assigned to the three favorable outcomes (benefits) to arrive at the benefits-barriers score.

The cues to action variable was obtained in a section of the questionnaire separate from that measuring the normative influences in the BIM and included references to TV, radio, and the print media as well as one's doctor, family, and friends. Respondents were asked to specify the media and persons used as information sources and to indicate whether this information was "for swine flu shots" (+2) at the one extreme of a five item scale, to "against" flu shots (-2) at the other. The sum of these responses was used to operationalize the cues to action construct.

The Drive Reduction Model. The drive arousal component of the DRM was measured with reference to the action null set, that of failing to get a swine flu shot. According to Wortzel (1976), this variable is defined as the probability of contracting swine flu in the absence of an inoculation times the anticipated severity of the problems associated with the disease. To obtain these measures, subjects were asked to first indicate their chances of catching swine flu if they did not get a shot (perceived susceptibility) and then to estimate their chances of incurring four negative outcomes if they did get flu (perceived severity), all on five item scales ranging from "no chance" to "certain". The four outcome scores were averaged and then multiplied by the susceptibility probability.

Drive reduction has been defined by Wortzel (1976) as the subjective probability of success associated with the preventive action less an assessment of costs of that action. These measures were taken from the probability of consequences scales discussed in the BIM. The probability of success variable was operationalized as the likelihood assigned to the one outcome "immunity from swine flu" while the costs of action were estimated by the sum of probabilities attached to the negative consequences of getting a swine flu shot. Drive reduction was then operationalized as the probability of immunity minus the average of the sum of costs.

Criterion. To obtain a continuous measure of behavioral intention, subjects were asked to indicate the "chances in 10" that they would get a swine flu shot on an eleven point scale ranging from "no chance (0) to "certain" (1).


Simple correlations were first calculated between the criterion and the antecedents suggested by the various models to test whether the predictor variables were significantly associated at the zero order level. The intention criterion was later regressed on all suggested antecedents simultaneously to determine if each variable would make a significant independent contribution to the explanation of variance in intention. Finally, the coefficients of determination generated by each model were compared in an effort to identify the model with the greatest predictive potential. These analyses were performed for the sample as a whole and then by student and resident subsamples to determine if differences in model component weights existed across the two subject groups.


Of the 3000 mailed surveys, 252 student and 148 resident questionnaires were returned as undeliverable for various reasons, the most common of which was "moved, no forwarding address." Of the remaining 2600 surveys, 335 usable student replies and 515 usable resident returns were received yielding response rates of 45% and 28% respectively. Of these respondents, twelve students and 46 residents indicated that they had already received the shot and were eliminated from the study resulting in a final sample size of 792 including 323 students and 469 residents. Fifty-one percent of the student respondents and 53% of the residents in the sample were male.

Correlations between the intention criterion and all variables suggested by the three models are shown in Table 2 for the total sample and the two subsamples. It is evident that all antecedent variables were significantly correlated with the criterion. Moreover, the magnitudes of these associations were higher in the BIM than in the other two models suggesting that the Fishbein approach may have greater validity. It is also interesting to note that, in those models where concepts are linked sequentially (the BIM and the HBM), the variables posited as having a direct effect on intention produced higher correlations than those suggested as having indirect effects only, a finding which supports the conceptual structure of these models.

Table 2 also shows that marked differences across subsamples were observed only for the cues to action construct in the HBM and for the drive reduction construct in the DRM. The data suggest that intentions of the community residents were more highly related to media and word-of-mouth influences than were students' intentions and that the drive reducing properties of the inoculation were somewhat more instrumental for resident intentions than for students. For the most part, however, the correlations obtained in the two samples for all models were surprisingly similar.

When the criterion was regressed on all variables in each model simultaneously, the results shown in Table 3 were obtained. It is immediately apparent that the BIM yielded coefficients of determination much larger than those obtained with the HBM which, in turn, produced higher coefficients than the DRM. Note further that all variables hypothesized to directly affect the intention criterion generated significant coefficients while those which are posited to have indirect effects (attitudinal and normative beliefs in the BIM; perceived susceptibility and severity in the HBM) resulted in attenuated coefficients when compared to the zero-order correlations in Table 2. This finding lends support to the sequential or intervening variable influences suggested by the BIM and the HBM.





Table 3 shows that somewhat higher coefficients of determination were obtained when using the resident sample data. This finding was particularly evident for the HBM and DRM models and may arise from the higher zero order correlations for certain of the variables. It would be speculative, however, to attribute meaningful differences in model structure across samples to the small differences observed in the regression coefficients. Such differences may be more easily explained by multicollinearity effects or range differences in the variables.


The results have shown that the BIM yielded much higher coefficients of determination than either of the other competitive models. As the reader may recall, Fishbein's conceptual framework differed in a number of important respects from other schemata and this may explain its predictive superiority. First, the BIM incorporates evaluative components at both the individual consequence level and at the summary attitude toward (evaluation of) the act level. This would appear to be critical because individual beliefs have little meaning unless one's affect toward a consequence or behavior is gauged (Fishbein and Ajzen, 1975). Thus, the implicit assumption in the HBM that all benefits are evaluated as equally good and all barriers or costs as equally bad is not necessarily accurate.

Second, the BIM explicitly includes the influences of referent persons and, in doing so, acknowledges that relevant others are instrumental in the formation of one's behavioral intentions. While the HBM does have a cues to action component which is somewhat isomorphic with Fishbein's normative beliefs, it does not provide for the various motivations to comply with those beliefs (which can be seen as another form of evaluation). Again, it may be inappropriate for proponents of the HBM to assume that all cues to action which support one's intention to engage in the preventive behavior are viewed as equally salient or credible by the consumer.

Finally, Fishbein's model includes two important intervening variables, attitude toward the act and subjective norm, between the summed attitudinal and normative beliefs and intention. These mediating constructs serve to "capture" the essence of the underlying beliefs and, if the results of this study are representative, suggest that individuals do, in fact, form summary judgments of positive and negative consequences and positive and negative social influences. These summary judgments, then, become the most direct influences (and therefore the best predictors) of intention. Neither the HBM nor the DRM contains these variables.

The difference in the coefficients of determination between the HBM over that of the DRM may be attributed to the greater number of variables in the former model. Specifically, the HBM explicitly recognizes external information sources on one's intention to engage in a behavior and is not as restrictive in terms of the benefits or favorable outcomes which can be considered in the model. Thus, proponents of drive models of preventive behavior may wish to add social and other influences in an effort to increase the explained variance in their criteria.

In summary, it appears that the inclusion of evaluative components, normative influences, and intervening summary measures enhance the degree to which a model aids in the explanation of preventive health intentions. The next section suggests some implications of these findings.


Profit and non-profit preventive health care marketers may find a number of the findings reported here of interest in improving their understanding of consumer health decisions. First, it is noteworthy that all three models investigated explained a statistically significant portion of the variance in behavioral intention suggesting that some of the antecedents of health intentions may be known. Consequently, practitioners and field researchers may wish to borrow from certain of the models to augment the decision making processes used by their organization.

Second, researchers concerned with the most complete determination of health care decisions (e.g., when an epidemic is imminent) may wish to combine concepts from two or more of the models. To illustrate, it appears that inclusion of one or more emotional components (e.g., perceived threat) in the BIM may increase its predictability beyond that obtained with the more rational components proposed by Fishbein. [Further analysis of the data used here provides only tenuous evidence for this suggestion. When intention was regressed on the four Fishbein components and perceived susceptibility, severity, and threat simultaneously, the coefficients of determination increased by 2% for the total sample, by 4% for the student sample, and 2% for the resident sample. This is not to say, however, that greater improvements would not be obtained in other research settings.] Alternatively, researchers with an HBM orientation may wish to include evaluations of the benefits and barriers as well as the individual's overall evaluation of the preventive health behavior.

Finally, it is clear that some variables in the models have more immediate impacts on one's intention decision. It follows that if a researcher or practitioner were interested primarily in prediction, a more parsimonious model consisting of direct effects only may suffice. In a similar manner, if one wished to attempt to change an existing behavioral intention, it may be more productive to work on the more immediate antecedents. For example, a person's overall subjective norm may be more amenable to influence in the BIM than would the orientations of the individual referent persons.


The conclusions drawn here are contingent on the methodology employed to test the models. In view of this, certain readers may wish to question the authors' interpretation of model structure as well as the operationalization procedures used. To the extent that these criticisms are valid, the findings remain tentative. We would hope that others would take issue with this presentation and test alternative models and methodologies. Hopefully, a discourse of this nature will result in beneficial refinements and a more complete understanding of preventive health care decisions.


Raymond A. Bauer and Donald F. Cox, "Rational Versus Emotional Communications: A New Approach," in L. Arons and M. May, eds., Television and Human Behavior (New York: Appleton-Century-Crofts, 1963).

Marshall H. Becker, "The Health Belief Model and Sick Role Behavior," Health Education Monographs, 2 (Winter, 1974), 409-19.

Marshall H. Becker and Lois A. Maiman, "Sociobehavioral Determinants of Compliance with Health and Medical Care Recommendations," Medical Care, 13 (January, 1975), 10-24.

Philip M. Boffey, "Anatomy of a Decision: How the Nation Declared War on Swine Flu," Science, 192 (May, 1976), 636-41.

Don E. Dulany, "Awareness, Rules, and Propositional Control: A Confrontation with S-R Behavior Theory," in D. Horton and T. Dixon, eds., Verbal Behavior and General Behavior Theory (New York: Prentice-Hall, 1968).

Martin Fishbein, "Attitude and the Prediction of Behavior," in Martin Fishbein, ed., Readings in Attitude Theory and Measurement (New York: Wiley, 1967).

Martin Fishbein and Icek Ajzen, Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research (Reading, Mass.: Addison-Wesley, 1975).

Martin Fishbein and Icek Ajzen, "Misconceptions about the Fishbein Model: Reflections on a Study by Songer-Nocks," Journal of Experimental Social Psychology, 12 (1976), 579-84.

James Jaccard, "A Theoretical Analysis of Selected Factors Important to Health Education Strategies," Health Education Monographs, 3 (Summer, 1975), 152-67.

Philip Kotler and Sidney J. Levy, "Broadening the Concept of Marketing," Journal of Marketing, 33 (January, 1969), 10-15.

Lois A. Maiman and Marshall H. Becker, "The Health Belief Model: Origins and Correlates in Psychological Theory," Health Education Monographs, 4 (Winter, 1974), 336-353.

lan M. Newman, ed., Consumer Behavior in the Health Marketplace (Lincoln, Nebr.: Nebraska Center for Health Education, University of Nebraska, 1976).

Irwin M. Rosenstock, "The Health Belief Model and Preventive Health Behavior," Health Education Monographs, 4 (Winter, 1974), 354-86.

Elaine Songer-Nocks, "Situational Factors Affecting the Weighting of Predictor Components in the Fishbein Model," Journal of Experimental Social Psychology, 12 (January, 1976), 56-69.

Jonathan Spivak, "Program to Inoculate All Americans for Flu Appears to be Ailing," Wall Street Journal, 56 (June 18, 1976), 1ff.

Lawrence H. Wortzel, "The Behavior of the Health Care Consumer: A Selective Review," in Beverlee B. Anderson, ed., Advances in Consumer Research, 3 (1976), 295-301.

Gerald Zaltman and Ilan Vertinsky, "Health Service Marketing: A Suggested Model," Journal of Marketing, 35 (July, 1971), 19-27.



Richard L. Oliver, University of Iowa
Philip K. Berger, University of Kentucky


NA - Advances in Consumer Research Volume 05 | 1978

Share Proceeding

Featured papers

See More


The Preference for Simultaneity: When Different Events Happen to Different People at the Same Time

Franklin Shaddy, University of Chicago, USA
Yanping Tu, University of Florida, USA
Ayelet Fishbach, University of Chicago, USA

Read More


When do More Options Produce Worse Choice?

Shannon Duncan, Columbia University, USA
Ulf Bockenholt, Northwestern University, USA
Eric J Johnson, Columbia University, USA

Read More


Thanks for Nothing: Expressing Gratitude Invites Exploitation by Competitors

Kelly Kiyeon Lee, Georgetown University, USA
Jeremy A. Yip, Georgetown University, USA
Cindy Chan, University of Toronto, Canada
Alison Wood Brooks, Harvard Business School, USA

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.