Personality and Consumer Behavior: Extensions

Citation:

Masao Nakanishi (1972) ,"Personality and Consumer Behavior: Extensions", in SV - Proceedings of the Third Annual Conference of the Association for Consumer Research, eds. M. Venkatesan, Chicago, IL : Association for Consumer Research, Pages: 61-66.

Proceedings of the Third Annual Conference of the Association for Consumer Research, 1972      Pages 61-66

PERSONALITY AND CONSUMER BEHAVIOR: EXTENSIONS

Masao Nakanishi, University of California at Los Angeles

In a recent review of past and current research in personality and consumer behavior. Kassarjian noted (1971):

Purchasing behavior, media choice, innovation, segmentation, fear, social influence, product choice, opinion leadership, risk taking, attitude change, and almost anything else one can think of have been linked to personality.

The four papers presented today by Landon (1972), Bither and Dolich (1972), Gardner (1972), and Kinnear, Taylor and Ahmed (1972) also reflect the special fascination toward the personality concept held by the researchers in consumer behavior. KassarJian claimed that the proportion of the total variance in the criterion variables explained by personality construct measures has been typically between 5 and 10 per cent in most studies. The papers presented in this session, with the exception of the Bither and Dolich paper, report similar lack of explAnatory power by personality measures.

The papers by Landon and Gardner both deal with the individual differences in need for achievement and their relationship with product choice. As he acknowledges, Gardner's study must be treated as exploratory due to the smallness of the sample imposed by the measurement technique--Thematic Apperception Test. In the Landon paper, the operational problems of measurement are somewhat alleviated by the use of paper and pencil tests, but his conclusions are in doubt because of major flaws in statistical analysis. The Kinnear, Taylor and Ahmed paper tried to explain ecologically constructive behavior on the basis of socio-economic and personality variables, but found instead that an attitudinal variable--perceived consumer effectiveness--was the most significant variable. Only in the technically competent paper of Bither and Dolich, do we find some positive results on the relationship between personality and store choice behavior.

Both KassarJian and Jacoby (1971) have attributed the lack of explanatory power to two common failings of past studies, i.e., 1) the slip-shod adaptation of instruments designed for clinical uses, and 2) the lack of specific logical hypotheses to be tested. Though fully in agreement with those conclusions, skepticism remains if significant improvements in future research findings will result merely by more careful selection of test instruments and more experimental research designs to test specific hypotheses. The low explanatory power of individuals' personality characteristics, doubtlessly a pervasive force in shaping one's behavior, may have stemmed in part from naive conceptualization of the relationships between personality and consumer behavior often held by researchers in the field.

One of the naive concepts is that somehow an individual's personality has direct effects on his product and brand choices, much in the classical analysis of variance sense. The papers by Landon and Gardner appear to share this concept. This viewpoint is prompted by some researcher's desire to find easy methods of identifying potential customers and/or segmenting the market, but its naivete has become more and more apparent with recent research in consumer decision making. He now know that an individual's decision to purchase a product or a brand is only a final result of a series of sub-decisions, such as decisions on store choice and choice criteria for each product category. Only recently a number of studies began exploring personality effects on the consumer decisions other than those of product/brand choice. Mark Alpert (1972), for example, investigated personality effects on choice criteria, while the Bither and Dolich paper in this session focused on the store selection decisions.

It may be argued that, if intermediate consumer decisions leading to the final product/brand choice are influenced by personality characteristics, then product/brand choice decisions must also be some function of personality. It would be difficult to refute this argument, but the functional relationships between personality and product/brand choice may not be so simple as to be amenable to analysis by standard statistical techniques which are often designed to measure linear relationships. The desire to find quick segmentation variables notwithstanding, it will be more fruitful to investigate the personality effects on each identifiable stage of the consumer decision making process separately. When we know more of personality effects on various stages, we may be able to synthesize the findings into a cohesive theory of personality effects on consumer behavior. Though studies by Alpert and Bither and Dolich cover only a small ground compared with the vast area of consumer behavior research, their effort should be considered as the beginning of a new and promising area of personality research .

Another naive view, often held unconsciously, is that of stationality of personality effects on individual's behavior. Most previous studies, including those reported in this session, may be called "one-shot" studies, the main objective of which is the search for behavioral correlates among personality construct measures over a sample of individuals taken at one point in time. Figure 1 (a) depicts the data set typically used. The behavioral data could be actual or hypothetical choices, intentions, or preferences for any alternatives, not necessarily products or brands. Those studies assume by design that an individual with a given personality configuration always behaves one way at different time points. This of course represents an extremely static view of human behavior and runs counter to what we know of its adaptive nature.

FIGURE 1

Kassarjian suggested a definition of personality as the "generalized patterns of response or modes of coping with the world." Those with the static viewpoint would interpret this as meaning the consistency of an individual's behavior over variety of situations, but personality may also be related to the manner the individual adjusts to the changes in his environment. In f&ct, it may be more correct to conceive personality as a moderator variable whose function is to moderate the effects of environmental changes on individual's behavior. This dynamic concept of personality effect have not been taken seriously by those who are interested in PersonalitY research.

Some researchers, for example, Brody and Cunningham (1968) and Fry (1971), tried to alleviate the lack of dynamism by incorporating situational and/or moderator variables in their research design and achieved a limited degree of success. Their objective was to isolate the main effects of personality on consumer behavior by explicitly adopting classical analysis of variance models. Figure 1 (b) gives the type of data set used by those studies. This research design implicitly assumes that moderator variables, such as perceived risk and specific self confidence, adequately summarizes feedbacks from past behavioral and situational events. However, as research in personality and persuasibility attests, personality may also affect over-time changes in moderator variables. It would, then, be more meaningful and potentially more powerful in predicting behavior if one can find out how moderator variables are formed and/or changed and how they interact with individual's personality. The one-shot research designs of Figure l tell us nothing of this type of interaction.

In order to investigate the full effects of personality on consumer behavior, it becomes necessary to obtain behavioral data and personality construct measures for each individual over different situations, accompanied by the data on situational and moderator variables. The data thus created would look as in Figure 2.

FIGURE 2

Analyzing data sets of this type would be a difficult undertaking, however. One might think of techniques analogous to a combination of cross-sectional and time series analysis in econometrics, but analysis would be made complex by interaction effects among predictor variables and lagged effects expected from the relationships between moderator variables and behavioral and situational variables. What is needed is an entirely new way of simultaneously analyzing those three types of variables.

An approach probably comes closest to the concept of personality as a moderator variable between behavior and environmental changes. Instead of combining personality construct measures and situational variables as predictors in a single model, one may construct a primary model which links behavioral data with situational and moderator variables at the individual level. This model is fitted to the data individual by individual and its parameters are estimated for each individual. Then secondary models may be constructed to specify the relationships between the estimated parameter set and personality construct measures. This two-step procedure is illustrated in Figure 3.

Since the two-step approach eventually links behavior as a function of personality, it may seem that it is no different from the conventional methods of analysis. But the difference is more than superficial. First, there is no need to restrict the primary and secondary models to naive linear forms. Fitting of non-linear models in the first stage followed by a multivariate analysis such as canonical correlation in the second stage is a perfectly permissible combination. It is conceivable to use as complex a model as the Howard-Sheth model in the first stage. Secondly, the two-step approach would become more attractive as the number of predictor variables (personality, situational, moderator) becomes large. The first stage becomes in fact a data reduction procedure, if the number of parameters to be estimated is small relative to the number of predictors.

FIGURE 3

The purpose of this discussion was to show how naive concepts of personality effects on consumer behavior tend to obscure the areas of research which are most interesting and rewarding. To a degree, the naivete may arise from the limitations imposed by conventional techniques of statistical analysis. Although it is a truism that only clear conceptualization of relationships among variables under study helps the researcher determine the data to be obtained and the analytical techniques to be used, it is also true that we tend to gravitate toward an easy habit of conceptualizing everything in terms of simplistic linear relationships. The more pervasive the effect of personality, the more critical it becomes to clarify the myriad of relationships among many layers of variables interacting with each other since such complex interactions may not reveal themselves as simple linear relationships. In order for us to be able to develop a meaningful theory of personality effects it will be necessary to break out of the mold into which naive conceptualization has forced us.

REFERENCES

Alpert, M. I. Personality and the Determinants of Product Choice. Journal of Marketing Research, 1972, 2, 89-92.

Bither, S. W. & Dolich, I. J. Personality as a Determinant Factor in Store Choice. Paper presented at the Third Annual Conference of the Association for Consumer Research, Chicago, Illinois, November 1972.

Brody, R. P. & Cunningham, S. M. Personality Variables and the Consumer Decision Process. Journal Of Marketing Research, 1968, 5, 50-57.

Fry, J. N. Personality Variables and Cigarette Brand Choice. Journal of Marketing Research, 1971, 8, 298-304.

Gardner, D. M. An Exploratory Investigation of Achievement Motivation Effects on Consumer Behavior. Paper presented at the Third Annual Conference of the Association for Consumer Research, Chicago, Illinois, November 1972.

Jacoby, J. Personality and Innovation Proneness. Journal of Marketing Research, 1971, 8, 244-247.

Kassarjian, H. H. Personality and Consumer Behavior: A Review. Journal of Marketing Research, 1971, 8, 409-418.

Kinnear, T., Taylor, J. R., & Ahmed, S. A. Socio-economic and Personality Correlates of Ecologically Constructive Purchasing Behavior. Paper presented at the Third Annual Conference of the Association for Consumer Research, Chicago, Illinois, November 1972.

Landon, E. L. A Sex-Role Explanation of Purchase Intention Differences of Consumers Who Are High and Low in Need for Achievement. Paper presented at the Third Annual Conference of the Association for Consumer Research, Chicago. Illinois, November 1972.

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