Effects of Prior Decision-Making, Demographics, and Psychographics on Marital Roles For Purchasing Durables

Arch G. Woodside, University of South Carolina
[ to cite ]:
Arch G. Woodside (1975) ,"Effects of Prior Decision-Making, Demographics, and Psychographics on Marital Roles For Purchasing Durables", in NA - Advances in Consumer Research Volume 02, eds. Mary Jane Schlinger, Ann Abor, MI : Association for Consumer Research, Pages: 81-92.

Advances in Consumer Research Volume 2, 1975      Pages 81-92


Arch G. Woodside, University of South Carolina

[Associate Professor of Business Administration and Program Director of Marketing.]

[Data collection assistance of Patty Richardson and Timothy Bice is gratefully acknowledged.]

Can marital roles be accurately predicted for who actually makes the purchases of automobiles and rugs/carpets from information on marital roles in prior stages in the decision-making process and from demographic and psychographic data? What variables are most associated with marital roles of actual product purchasing? Results of this study of 200 husbands and 200 wives suggest that patterns for developing family types in consumer decision-making do exist.

Ferber (1973) has made a strong case for the greater need for product purchase models compared with brand purchase models:

. . . marketing researchers have been placing too much emphasis on using concepts from attitude theory to explain brand purchases and too little to explain product purchases. After all, the key problem is to anticipate sales of the product, which for many products may vary substantially from one year to another, whereas for most products brand shares vary little from one year or quarter to another.

This greater need for theoretical and empirical analysis of product decision-making by consumers may become apparent with the arrival of the 1974-1975 recession. Mid-October new car sales in 1974 were the lowest since 1964 and nearly 30% below 1973 sales (UPI, 1974), and, the 1974 slackening demand for other consumer goods suggests the basic question for consumer research is how consumers enter, pass through, and actually make product purchase decisions.

Fortunately, the study of family buying decisions has been concentrated on product purchase behavior (Hempel, 1974; Davis, 1970: Ferber and Lee, 1974: Davis and Rigaux, 1974; Sheth, 1974) which has produced substantial insights into the relative roles of husbands and wives in buying particular products and the effects of life cycle. social class, employment status of wife, social networks, and prior decision-making on these buying roles (Sheth, 1974).

The Engel, Kollat, and Blackwell (1973, ?. 58) model of consumer behavior is the most related theoretical statement to extant research into family buying decisions and product purchase behavior. Both the Engel, Kollat, and Blackwell model and family buying research concentrate on the stages in cognitive decision processes of product purchases: problem-recognition, search for information, evaluation, store choice and shopping decisions, purchasing, and postpurchase processes. Other models and research of consumer behavior have concentrated on the relations between perceptions, attitude, intentions and purchases of brands (Howard and Sheth, 1969: Hansen, 1973).

The focus of the present study is the effects of prior decision-making, demographics, and psychographics on husband and wife roles for purchasing durables. Can marital roles be accurately predicted for who actually makes the purchases of automobiles and rugs/carpets from information on marital roles in prior stages in the decision process and from demographic and psychographic data? What variables are most associated with marital roles of actual product purchasing? What are the differences in these salient variables for husband, wife, and syncratic purchases?

A basis for developing family types based upon the effects of prior decision-making and demographic and psychographic information would exist given substantially accurate predictions of marital roles in product purchasing. Davis (1970) has found that little basis for developing family types exists from patterns of husband-wife influence across several decisions when the decisions are not considered in independent and dependent relationships, e.g., where to buy not made dependent on when to buy. However, the relative influence of husbands and wives in making actual purchases may indeed be dependent upon prior decision-making (and demographic and psychographic variables) which would be suggested by the Engel. Kollat. and Blackwell (1973) model.


The data were obtained from a cross-sectional survey of 200 families from three housing subdivisions in the Columbia, South Carolina, metropolitan area in 1972. Families were selected from a random sample of street tracts of single family dwellings with two families surveyed on each street. The social classes (defined by occupation and income ratings) were distinctly different for the families from the three subdivisions: u?per-lower, lower-middle, and upper-middle social divisions.

Husbands and wives were interviewed separately in the 200 families. Interviewers were male and female graduate students. One interviewer went into each house separately. Families selected for the study were contacted prior to the interview by letter and telephone.

The survey instrument consisted of two parts. Husbands and wives answered a series of questions on the relative influence of each spouse for eight products: automobiles, lawnmowers, automatic washing machines, beer, rugs/carpets, cheese, television sets, and gardening supplies. Secondly, the husbands and wives completed a demographic and psychographic instrument which included questions on family life cycle, occupation of family head, wife employment, education, income, number of visits by friends in neighborhood, popularity, conservatism, club activities, advertising attitudes, opinion leadership, and other psychographic questions. The psychographic questions were selected from 300 factor analyzed questions developed by researchers at Purdue University (Tigert, 1969). Typically, four questions were included from each factor within the survey. Responses were true and false for each item and the factor was scored from 1 to 4.

The responses for each spouse on relative influence were scored by assigning a value of 1 when the husband's role was perceived as dominant, 2 for joint decisions, and 3 when the wife's role was dominant.

The decision-making questions for the eight products included questions on who brought up the idea of purchasing the product, discussion of the purchase with friends, neighbors, relatives, obtaining information from mass media, obtaining information from stores (dealers), style or type decisions, who visited stores or dealer showrooms, who selected the specific retail outlet to purchase, who made the actual purchase, and who experienced dissatisfaction, if anyone.

The assumptions and limitations of the direct questioning approach of family roles in decision-making have been discussed elsewhere (Kenkel, 1961; Davis and Rigaux, 1974). In particular, the question of ability to recall with accuracy how the influence was distributed in some past decision-making session should be empirically studied. However, the direct questioning approach does permit a richness in data analysis since a large number of decision stages and products can be included in the same study and the validity of the measurement of perceived influence of husbands and wives separately has been supported (Davis, 1971). Engel, Kollat, and Blackwell (1973, p. 203) offer the following conclusion on measuring family role structure:

At the present time there does not appear to be any research design that can overcome the problems involved in identifying roles. The best interim approach seems to be to use direct questions about specific decisions and activities at each stage in the decision-making process for the marketer's product or service.

The following analyses of patterns of relative influence of husbands and wives are for two durable goods purchases, an automobile and rugs/carpets. Both of these products involve substantial financial outlays, extended periods of ownership, social importance, use by several family members, and most likely different patterns of husband and wife influence. Prior research (Cunningham and Green, 1974- Davis, 1970) has found husbands to be more often dominant than wives for some automobile decisions, e.g., what make of automobile to buy. Wives have been found (Jaffee Associates, 1965) to be more often dominant than husbands for some carpets/rugs decisions, e.g., who first brought up the idea and who actually purchased the product.


Substantial variability between decisions and between products in husband-wife roles can be found in the distribution of husbands' and wives' responses as shown in Table 1. Both husbands and wives responses show high percentages of husband-dominant families for bringing-up the idea to purchase an automobile (70.6% of both husband and wife responses) and in actually purchasing the automobile (67% and 66.5% of the respective husband and wife responses). Some of the other decisions and purchase activities show decreases in husband-dominance and increases in syncratic behavior for automobile purchases, e.g., 66.7% of the husbands and 67.6% of the wives reported that visiting dealer showrooms to be a joint activity.

More husbands and wives reported that the wife had greater dominance in rugs/carpets decisions compared with automobile decisions. The percentage of wife-dominant families decreases from 60.1 for bringing-up the idea to purchase to 26.8% for actually purchasing among husband responses for rugs/carpets. Wife responses are similar in percent compared with husband responses. Higher percentages of husband-dominance is shown in Table 1 for making the actual purchase than for any other decision or activity, over 20% of both husbands and wives reported that the husband alone made the actual purchase for rugs/ carpets.

High levels of agreement in relative influence among husbands and wives are shown in Table 1 (for aggregated data). The rank orders of percentages of families with husband-dominant or wife-dominant are nearly identical across the 11 decisions for automobiles and rugs/carpets. This finding has been previously observed for automobiles and furniture (Davis, 1970), and automatic washing machines and television sets (Woodside, 1972). Therefore, similar conclusions on relative influence would be likely reached by analyzing either aggregated husband or wife responses; at the same time, some substantial differences within families may exist for husband and wife responses.



Patterns of relative husband-wife influence across decisions and demographic-psychographic information for predicting who actually purchased the two products were analyzed through stepwise multiple discriminant analysis (MDA). Husbands and wives were separately grouped by husband, joint, or wife-dominance for actually purchasing automobiles and rugs/carpets. The independent variables included 10 prior decision-making variables as shown in Table 1 and 26 demographic-psychographic variables.

The average responses and standard deviations of four prior decision-making variables and one psychographic variable included in the first five variables to enter the multiple discriminant functions for automobiles among the husband responses are shown in Table 2. A total of 82.7% of the 191 husbands were correctly classified into husband or joint decision-making in actually purchasing automobiles. k wife-dominant group was not included in the analysis since so few wives dominated the actual purchase decision. Results are shown only for 5 variables in Table 2 since additional variables did not substantially increase the level of correctly classifying respondents.

The standardized betas shown in Table 2 indicate the most salient variables in discriminating relative husband-wife influence. The likelihood of jointly making the actual purchase of an automobile increases if friends (and neighbors or relatives) are jointly consulted prior to purchase. Jointly visiting showrooms and choosing a specific dealer increases the likelihood of jointly making the actual decision; the same finding holds for jointly deciding how much to spend for the automobile. Classification as husband-dominant in actually purchasing an automobile is more likely, the more favorable the attitude toward private brands.

Decisions on bringing-up the idea to purchase, style, size, brand, and consulting mass media were substantially less important than the variables discussed in correctly classifying husbands by relative influence in actually purchasing automobiles.

The analysis of the wife responses for automobiles is shown in Table 3. A total of 82.9% of the wives were correctly classified into husband-dominant or joint decision-making in making the actual purchase decision by 5 variables in the multiple discriminate functions.

High income level increased the likelihood of classification into husband-dominance in actually purchasing automobiles while high conservatism increased the likelihood of joint decision-making for the wife responses. Relative influences of husbands and wives in deciding how much to spend, visiting showrooms, and choosing a particular dealer were significantly related to the relative influence of husbands and wives in actually purchasing automobiles, as analyzed from the wife responses.

The percentages of correctly classified respondents for automobiles were substantially higher than would be expected by either the proportional chance criterion (57%) or the maximum chance criterion (68%). and these percent accuracies may be assumed to be impressive (Morrison, 1969). However, the results are somewhat biased upward since all the data were used to estimate the discriminate coefficients (Frank, Massy, and Morrison, 1965).





Data in Tables 4 and 5 are the means, standard deviations, standardized betas, and confusion matrixes for husband and wife responses in classifying families by relative influence in actually purchasing rugs/carpets. The relative influences on the size of the rug/carpet, visiting stores, and choosing the specific outlet were important in classifying families for both husband and wife responses. The greater the wife involvement in these decisions except for determining sizes, the greater the likelihood of wife-dominance in the actual purchase decision. Neighborhood popularity and consulting friends were significant in classifying relative influence in actually purchasing rugs/carpets for the husband responses, while the employment status of the wife and the relative influence on how much to spend were important in classifying relative influence in actually purchasing rugs/carpets for the wife responses.

Wife employment was entered into the functions as a dummy variable: 1 for wife as a full-time housewife and 2 for wife employed outside the home. Results in Table 5 would suggest that wife employment outside the home increases the likelihood that the wife would be involved in making the actual purchase of rugs/carpets.

The percentages of correct classifications for relative influence in actually purchasing rugs/carpets are substantial for joint and wife-dominance and less impressive for husband-dominance as shown in Tables 4 and 5.


The findings support the hypothesis that some basis of developing family types does exist through the analysis of relative influence in prior decision-making and demographic-psychographic data. The predominance of relative influence in prior decision-making in predicting husband and wife roles in actual purchases of automobiles and rugs/carpets supports the cognitive process model of purchase behavior by Engel, Kollat, and Blackwell (1973).

Demographic and psychographic variables do appear to be associated in a more limited role with actual purchase behavior compared with prior decision-making variables. Demographic and psychographic data may be more likely associated with the relative influences of husbands and wives in initiating the idea of purchasing rather than the purchase itself. This hypothesis should be explored in future research.






Cunningham, C.M. & Green, R.T. Purchasing roles in the U.S. family, 1955 and 1973. Journal of Marketing, 1974, 38, 61-64.

Davis, H.L. Dimensions of marital roles in consumer decision making. Journal of Marketing Research, 1970, 7, 168-177.

Davis, H.L. Measurement of husband-wife influence in consumer purchase decisions. Journal of Marketing Research, 1971, 8, 305-312.

Davis, H.L. & Rigaux, B.P. Perception of marital roles in decision processes. Journal of Consumer Research, 1974, 1, 51-62.

Engel, J.F., Kollat, D.T. & Blackwell, R.D. Consumer behavior. (2nd ed.) New York: Holt, Rinehart, & Winston, 1973.

Ferber, R. Applications of behavioral theories to the study of family marketing behavior. Faculty working paper no. 125, University of Illinois, 1973.

Ferber, R. & Lee, L.C. Husband-wife influence in family purchasing behavior. Journal of Consumer Research, 1974, 1, 43-50.

Frank, R.E., Massy, W.F., & Morrison, D.G. Bias in multiple discriminant analysis. Journal of Marketing Research, 1965, 2, 250-258.

Hansen, F. Consumer choice behavior: A cognitive theory. New York: The Free Press, 1972.

Hempel, D.J. Family buying decisions: A cross cultural Perspective. Journal of Marketing Research, 1974, 11, 295-302.

Jaffe Associates. A pilot study of the roles of husbands and wives in purchasing decisions. New York: Life Magazine, 1965.

Kenkel, W.F. Family interaction in decision-making on spending. In N.N. Foote (Ed.), Household decision-making. Volume 4. New York: New York University Press, 1961.

Morrison, D.G. On the interpretation of discriminant analysis. Journal of Marketing Research, 1969, 6, 156-163.

Sheth, J.N. A theory of family buying decisions. In J.N. Sheth (Ed.), Models of buyer behavior. New York: Harper and Row, 1974.

Tigert, D. A psychographic profile of magazine audiences: An investigation of a media's climate. Unpublished paper presented to the Consumer Behavior Workshop. Columbus: Ohio State University. 1969.

UPI (United Press International). Automakers say sales at 10-year October low. Columbia, South Carolina: The State, October 24, 1974.

Woodside, A.G. Dominance and conflict in family purchasing decisions. Proceedings, Association for Consumer Research, 1972.