Husband and Wife Purchase Decision-Making Roles: Agreed, Presumed, Conceded, and Disputed

Alvin C. Burns, Louisiana State University
ABSTRACT - The author criticizes methods of measuring husband-wife purchase role structure and offers an extended conceptualization incorporating modes of role consensus and nonconsensus. Use of the method on aggregate data from a convenience sample reveals husband role overestimation and wife role underestimation across decision categories; unique predilections of couples are suggested through individual analysis.
[ to cite ]:
Alvin C. Burns (1977) ,"Husband and Wife Purchase Decision-Making Roles: Agreed, Presumed, Conceded, and Disputed", in NA - Advances in Consumer Research Volume 04, eds. William D. Perreault, Jr., Atlanta, GA : Association for Consumer Research, Pages: 50-55.

Advances in Consumer Research Volume 4, 1977   Pages 50-55


Alvin C. Burns, Louisiana State University


The author criticizes methods of measuring husband-wife purchase role structure and offers an extended conceptualization incorporating modes of role consensus and nonconsensus. Use of the method on aggregate data from a convenience sample reveals husband role overestimation and wife role underestimation across decision categories; unique predilections of couples are suggested through individual analysis.


A customary approach to study of husband and wife purchase decision role structure, influence distribution and/or authority is to ask each spouse about relative participation in the process. These separate assessments and then combined or compared in some manner to determine the nature of the decision-making process. No less than three different methods of combining or comparing husband and wife responses are evident in the literature. Each one holds a danger of distorting responses into a false characterization of the actual role structure. This paper addresses the methodological problem of role structure measurement and suggests a more precise means of reporting spouses' responses.

The first of these three methods involves securing spouses' evaluations of the character of the decision with response categories such as, "husband more than wife" or "wife more than husband." Several variations exist. These responses are aggregated by spousal group and the husband group central tendency is compared to the wife group central tendency to draw generalizations about syncratic or spouse-dominant product decisions. This method typifies many early husband-wife role structure studies and is still used in contemporary studies (see Woodside, 1975, for example). Criticism of this method has concentrated on its failure to represent individual couples. Aggregate analysis, it is pointed out, does not consider whether or not a person agrees or disagrees with his/her spouse. If fifty percent of the husbands designate a decision to he Joint and fifty percent of the couples agree that the decision should be jointly decided, in the extreme case, it is possible that none of the couples agree that it is a joint decision. Consequently, husband group to wife comparisons yield defensible generalizations of role structure only when both groups indicate high percentages on the same category and the probability of matched agreement is therefore high.

An alternative and more common method is to compare each husband's responses with his wife's responses. This procedure overcomes the aggregate group analysis problem and allows the researcher to discuss individual couple agreements and disagreements. Heer (1962), Safilios-Rothschild (1969), Granbois and Willett (1970, Davis (1970), Davis and Rigaux (1974), and Shuptrine and Samuelson (1976) are examples of this method. The results afford a more appropriate determination of joint and spouse-dominant decision areas inasmuch as there is the assurance that matched spouses agree. Misperceptions of individual spouses are eliminated. Nevertheless, there has been reported a substantial percentage of couples who do not agree. In fact, informal analysis on the above studies suggests that roughly one-third of the couples disagree, on the average. Certain decisions hold the potential for greater disagreement than others. Dunsing and Hafstrom (1975) have observed that "... when the responses of wives are compared with those of their own husbands, the results vary considerably depending on what decisions are being investigated." Consequently, while this method is better than aggregate group analysis, it does not account for a considerable proportion of the couples under study.

The third method of assessing role structure requires that spouses' responses be assigned a position on an interval scale. Traditionally, a three or a five-point scale has been used, although extended scales exist. Davis (1970), for example, used a five-point scale measure of "who decided" (husband decided=1; husband more influence than wife=2; etc.) and Davis and Rigaux (1974) incorporated a three-point scale. Implicitly, this procedure assumes that a continuum of husband-joint-wife participation underlies the decision making process and the assigned values (e.g., 1,2.3) measure role structure accurately. The method is clearly appealing as the equal interval assumption permits more elegant statistical analyses than otherwise is the case. Two criticisms may be voiced against this procedure. In the first place, the scale values are arbitrary and entirely inappropriate when spouses have responded to categories. (It is somewhat more defensible when they have indicated intensity of influence). The second criticism involves the results of the analysis. If a husband indicates "husband" (=1) and his wife indicates "wife" (=3) the couple's average is a joint decision (=2).


Some attention has been devoted to differences between spouses role assessments. Heer (1962), Safilis-Rothschild (1969), and Granbois and Willett (1970), for instance, discuss over- and under-estimation of influence. Similarly, Davis and Rigaux (1974) have described the presence of nonconsensus as "vanity" and "modesty" evident in spouses estimates of their influence for each of various product categories at stages in the decision process. They note that nonconsensus is a form of perceptual bias and urge that it he investigated with regard to magnitude and skewness.

With this advice in mind, the author has come to realize that a number of different types of nonconsensus are distinguishable when spouses' responses are compared. The following paradigm (Figure 1) shows how nine modes of consensus and nonconsensus can be identified when three response categories--"husband", "joint", and "wife"--are used. The three cells, Husband-Husband, Joint-Joint, and Wife-Wife define the consensus diagonal. The remaining six cells constitute variations of nonconsensus. At least four qualitatively different types of husband-wife decision-making role structure nonconsensus are apparent: disputed, presumed, conceded, and dually conceded. Each of these warrants discussion.



Agreed Decision Making Roles

When spouses indicate the same response with regard to role structure in the decision making process, i.e., that is either a husband decision, a wife decision, or a joint decision, the perceptions or specifications may be termed "agreed." It is the analysis of the consensus diagonal which has been customary from those researchers who have opted to match husband's with the wife's responses. Decision areas are denoted as husband-dominant, wife-dominant, or Joint depending on the relative percent of couples determined for each case.

Conceded Decision-Making Role

Two cells comprise instances in which the decision-making role is conceded. These are instances in which one spouse believes that role is Jointly held, but the other spouse believes that the role is taken by the first. The role is therefore conceded by one spouse to the other. [In this paper, nonconsensus is defined in terms of the speaker. For example, "wife concession" is to be interpreted as "wife concedes the role to the husband."] In reality, the conceding spouse underestimates his or her role, crediting the other spouse with more influence, participation, or authority than that spouse is willing to indicate.

Presumed Decision-Making Role

There are two cells in which one spouse assumes he or she takes the role, but the other spouse believes the role was jointly held. Thus, the presuming spouse overestimates his or her participation, influence or authority relative to the partner's assessment,

Disputed Decision-Making Role

Disputed roles result when spouses give opposing responses wherein the husband believes he takes the role and the wife believes she takes the role. Both spouses greatly overestimate their own roles in the decision relative to the other spouse's estimate.

Dually Conceded Role

Each spouse may credit the other spouse with the role: the husband attributes the role to the wife and the wife attributes it to the husband. In this case, both spouses greatly underestimate their own roles in the decision relative to the other spouse's estimate.

The paradigm can be applied to situations in which spouses respond to an intensity scale (e.g., a five-point scale ranging from "wife more than husband" to "husband more than wife"). Rather than cells, the intersections of paired responses define regions, as for example in Figure 2. With the use of scale values, it is necessary to specify the boundaries of the regions defining, for instance, the point at which the spouses' responses do not agree, are not opposing, etc.



Adoption of this conceptualization holds the potential for more comprehensive descriptions of husband and wife role structure at both the macroanalytic and the micro-analytic level, Suppose, for example, that husbands and wives responded to the categories in Figure 1 to describe the role structure of various purchase decisions. The percentage of couples falling in each cell would describe the distribution of husbands' and wives' matched responses for a given decision at a particular stage in the decision-making process. This approach would be especially valuable in cases where husband-wife agreement is low but a relatively large number fall in a specific nonconsensus cell, Alternatively, even when a majority of agreement is found in any consensus cell, greater insight would be afforded were it determined that a certain nonconsensus cell contained a major proportion of the remaining couples. A particular decision might be described "wife-agreed, wife-presumed" for example.

At the microanalytic level, greater understanding of the nature of decision-making strategies and processes could eventuate with the knowledge that a couple or group of couples falls into a certain cell or region. Implicitly, each of the types of nonconsensus denotes different outcomes and tactics. Some, such as the conceded role, suggest that one spouse will assume a more passive attitude in the decision than would be the case for the presumed role. The disputed and dually conceded role cells also suggest the nature of the decision process. Alternatively, the conceptualization could be applied to a single couple across several decisions as a means of analyzing perceptual bias.


The remainder of this paper reports results and conclusions from different analyses using the extended conceptualization of husband and wife decision-making role structure. Specifically, the questions addressed here are:

1. What are the relationships among consensus and non-consensus? In a group of husbands and wives, for example, can we expect wife presumption nonconsensus to be associated with husband concession nonconsensus as intuition suggests? Are systematic biases in evidence?

2. Are there differences in the relationships between the cells in the cases of husband-dominant, wife-dominant, or joint decisions? Is agreed husband dominance, for instance, more likely to be paired with wife concession than is agreed wife dominance to be paired with husband concession? A finding of this nature would help to evaluate previous and future role structure results by suggesting the character of the "other" category.

3. Are there differences between couples in their predilections for certain types of consensus or non-consensus? Investigation of individual differences between couples might help to explain the variability typically reported in husband and wife studies.

The data used for this study was part of a larger study on husband and wife role structure focusing on the final decision stage of the decision-making process. In the study, spouses responded separately to the question, "If you and your spouse disagreed on ________, who has the right to make the final decision?" Response categories we-re "husband", "wife", or "together" (joint). Spouses were not allowed to compare answers during the study. A convenience sample of 101 married couples was instructed to give responses to each of thirty-nine product decision areas, most of which were related to products and services requiring large financial outlays or commitments. The average couple tended to be middle-aged, well-educated, and financially secure. More complete demographic characteristics and information about the larger study are reported in Burns (1976).


Preliminary analysis of couples' responses took the form of cross-classification tables for each decision area, identical to that shown in Table 1 which illustrates the results for the decision "brand of a new television." From this table, it is apparent that while the majority of couples are in agreement (52% for joint, 12% for husband, and 1% for wife), over one third (35%) exhibit some form of nonconsensus. Inspection of the nonconsensus cells reveals that husband presumption and wife concession (14% and 17% respectively) are the predominant modes.



Table 2 offers the results for all thirty-nine product decision cross-classifications in tabular form. In the interest of convenience and for future reference they are arranged in three groups based on the majority of agreed couples: husband-dominant, wife-dominant, and joint decisions. Row totals are not equal to 1.00 in some cases due to rounding. Across all product decisions, an average of over one-third of the couples in this sample demonstrates nonconsensus; however, it is noteworthy that disputes and dual concessions are almost nonexistent in this sample. The majority of non-consensus appears in the forms of presumption and concession. Attention will be concentrated on these two modes for the remainder of the paper as data on the other two is insufficient.

Inasmuch as the first two research questions concern aggregate analysis, the proportions in Table 2 served as data in the search for relationships between the consensus and the nonconsensus modes. Pearson product moment correlations were computed for the total set of decisions, and the three subsets classified as husband-dominant, wife-dominant, and joint decisions in Table 2. The results for the total set are contained in Table 3; Table 4 compares the three types of decisions. This method of analysis is not entirely proper; nonetheless, it is defensible given that the concern is with relative differences.

A number of the correlations in Table 3 were found to be significantly different from zero correlation at the .01 level of significance. Consensus on husband influence tends to be negatively associated with consensus that the decision will be resolved jointly, also wife presumption, and husband concession. While the values of the correlations are low, their directions support the statement that husband role overestimation and wife role underestimation coincide with agreement on the husband's dominant role. Consensus on wife influence tends to be negatively associated with agreed joint resolution, husband presumption and wife concession. Moreover, there appears to be a greater likelihood for the husband to overestimate his role relative to the wife's estimate of his role than in the previous instance. Consensus on joint decisions tends to be negatively associated with wife presumption, and agreed husband or wife dominance. The remaining significant correlations between nonconsensus modes reveal patterns which are intuitive (e.g., husband presumption is negatively associated with wife assumption). An interesting finding does occur in the form of high correlation between husband presumption and wife concession which does not appear to be reciprocated to the same degree by the correlation between wife presumption and husband concession. A reasonable conclusion from this stage of aggregate analysis is that there exists a general tendency in husbands to overestimate their roles and in wives to underestimate theirs.

Table 4 illustrates that the distributions of nonconsensus loadings vary with husband-wife role structure categorizations based on the majority of consensus. Interpretation is not straightforward due to the use of proportions, and reference to Table 2 is helpful. In the case of husband-dominant decisions, the results indicate that as less agreement occurs on the husband's dominant role, wife concession nonconsensus and husband presumption consensus increases (the latter is significant at the .05 level). In the instance of wife-dominant decisions, the strongest tendency is toward agreement on joint participation as wife agreement declines. Also, there is a greater probability of husband concession than wife presumption. The two sets of results imply that many wives underestimate their roles and many husbands indulge in overestimation. Joint decisions results in Table 4 suggest fragmentation of nonconsensus with no discernible pattern. This result is probably attributable to the fact that the joint decisions with lower proportions of couples in the joint agreement cell usually have agreement on a single spouse rather than both; that is, some are "joint and wife" and some are "joint and husband". The nonconsensus cells verify this observation: wife presumption and husband concession increases directly with greater wife agreement, and wife concession and husband presumption increases less directly with greater agreed husband role dominance. Apparently, agreed joint participation decisions which also have some couples who agree on wife role dominance signal the presence of two other types of couples: those in which the wife presumes the role is hers and those in which the husband concedes the role to her. A similar but less strong pattern of nonconsensus distribution occurs for agreed joint decisions paired with couples agreeing on husband role dominance.

An analysis of individual couples was undertaken to ascertain the associations between consensus and nonconsensus at this level. In this phase of the analysis, the proportions of decisions falling in each cell for each couple were computed and used as the data set. Correlations were computed across all the decisions, but the correlation matrix is not reproduced here inasmuch as the vast majority of correlations failed to demonstrate statistical significance from no correlation at the .01 level. Furthermore, those coefficients which were significant were of small magnitude. The highest correlation, for example, was -.54 between husband presumption and wife concession. It does not appear that systematic relationships exist at this level of analysis.



Further analysis of individual differences was undertaken to determine predilections of couples for unusually large proportions of consensus or nonconsensus. Presentation of the frequency distributions is cumbersome and subsequently omitted. Means and ranges are found in Table 5. The minimum was 0.0 in all cases except agreed joint decisions where it was .05. While it is not readily apparent from this table, the results did suggest that particular couples exhibit predilections toward particular dispositions. For example, almost 20 percent of the couples demonstrated joint agreement on a majority of the total set of decisions. Simultaneously considering the high correlations determined in aggregate analysis, the low and nonsignificant correlations in individual analysis, and the frequency distribution results, it is reasonable to conclude that a considerable number of couples do demonstrate predilections for a single form of consensus or nonconsensus when their responses are compared.








The results indicate that the use of the extended conceptualization is of value at both the macroanalytic and microanalytic levels of husband and wife purchase decision role structure. Furthermore, its use with this convenience sample has revealed some interesting patterns of spouses' over- and underestimation of their own and the other's participation in the actual decision stage of the process.

At the aggregate analysis level, there is a general tendency across all decisions for husbands of some couples to overestimate their influence, participation, or authority and a tendency in other couples for wives to underestimate their roles. This finding implies that agreed husband-dominance probably contains an element of perceptual bias on the parts of both spouses. Agreed wife-dominant decisions, on the other hand, are probably more faithful to an objective measure of the process, if one were available. This implication casts doubt on the precise nature of husband and wife role structure when both spouses agree that the husband has taken or will take the major role. Some degree of joint participation is misperceived as husband dominance. Similarly, some degree of wife dominance is misperceived as joint role structure.

Further insight into this condition is apparent in the analysis of husband-dominant, wife dominant, and joint decisions. The character of nonconcessions is most evident in those cases where only a slight majority of the couples agree on role participation. In product decisions classified as husband-dominant, for example, one can also expect to find with equal probability two types of couples--those in which the husband presumes the role and those in which the wife concedes the role. In the case of wife-dominant decision areas, a number of couples exhibit husband concession (i.e., wife underestimation of her role.) Another set of couples agrees the decision is joint, a result which also could be interpreted as husband overestimation and wife underestimation of their own roles. The pattern of nonconsensus in joint decisions lends further support to this general conclusion.

Finally, individual analysis suggests that some couples have inclinations with respect to various types of consensus and nonconsensus on role structure. A number of couples agree on joint role structure for many decisions, another group tends to exhibit presumption and another group tends to show wife concession. Conceivably, these categorizations may be most useful in the description of role structure at the individual couple level, for knowledge of a couple's disposition could enhance prediction of the outcome and tactics of preference conflict resolution.

The extended conceptualization has proven to be a worthwhile method of describing husband and wife role structure in this study. It has revealed general over- and underestimation biases in these husbands and wives, and has pointed out the nature of nonconsensus for husband-dominant and wife-dominant decisions. Its use has also been valuable in the analysis of individual differences. These positive results and the potential benefits of an expanded delineation of husband and wife role structure lead the author to recommend its use in future studies, particularly in situations where compared structure estimates do not agree.


A. G. Burns, "Spousal Involvement and Empathy in Jointly-Resolved and Authoritatively-Resolved Purchase Subdecisions." in Beverlee B. Anderson, ed., Advances in Consumer Research, Vol.3. Proceedings of the 6th Annual Conference of the Association for Consumer Research, Cincinnati: Association for Consumer Research, 1976, 199-207.

H. L. Davis, "Dimensions of Marital Roles in Consumer Decision Making." Journal of Marketing Research, 1970, 7, 168-177.

H. L. Davis and Rigaux, B. P., "Perceptions of Marital Roles in Decision Processes," Journal of Consumer Research, 1974, 1, (2), 51-62.

M. M. Dunsing and Hafstrom, J. L., "Methodological Considerations in Family Decision-Making Studies," in M. J. Schlinger, ed., Advances in Consumer Research, Vol.2. Proceedings of the 5th Annual Conference of the Association for Consumer Research, Chicago: Association for Consumer Research, 1975, 103-111.

D. h. Granbois and Willett, R. P., "Equivalence of Family Role Measures Based on Husband and Wife Data," Journal of Marriage and the Family, 1970, 32, 68-72.

D. M. Heer, "Husband and Wife Perceptions of Family Power Structure," Marriage and Family Living, 1962, 24, 65-67.

F. K. Shuptrine and Samuelson, G., "Dimensions of Marital Roles in Consumer Decision Making: Revisited," Journal of Marketing Research, 1976, 13, 87-91.

C. Safilios-Rothschild, "The Study of Family Power Structure: A Review 1960-1969," Journal of Marriage and the Family, 1970, 32, 539-552.

A. G. Woodside, "Effects of Prior Decision-Making, Demographics, and Psychographics on Marital Roles for Purchasing Durables," in M. J. Schlinger, ed., Advances in Consumer Research, Vol.2. Proceedings of the 5th Annual Conference of the Association for Consumer Research, Chicago: Association for Consumer Research, 1975, 81-91.