Role Transferal in the Household: a Conceptual Model and Partial Test

Mary Lou Roberts, Boston University
Lawrence H. Wortzel, Boston University
ABSTRACT - Changing life styles, including increasing labor force participation by women, have generated substantial interest in the subject of household task performance by males. This paper presents a conceptual model of household role transfer based on role transfer models from the organizational theory literature. It concludes, on the basis of a limited empirical test, that the model is worthy of further empirical testing.
[ to cite ]:
Mary Lou Roberts and Lawrence H. Wortzel (1982) ,"Role Transferal in the Household: a Conceptual Model and Partial Test", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 261-266.

Advances in Consumer Research Volume 9, 1982      Pages 261-266


Mary Lou Roberts, Boston University

Lawrence H. Wortzel, Boston University

[The authors wish to thank Needham, Harper & Steers Advertising Inc. for providing access to its 1980 Life Style Study data which were used in this project.]


Changing life styles, including increasing labor force participation by women, have generated substantial interest in the subject of household task performance by males. This paper presents a conceptual model of household role transfer based on role transfer models from the organizational theory literature. It concludes, on the basis of a limited empirical test, that the model is worthy of further empirical testing.

Over the past few years considerable interest has been shown in the phenomenon of household task performance by males. Both academic (Hendrix, Kinnear and Taylor 1979; Roberts and Wortzel 1980) and commercial (Cunningham and Walsh 1980, Benton and Bowles 1980, Time undated) studies have suggested that increasing male performance of household tasks is an important trend which has numerous implications for the marketing of food and home care products. The usual assumption is that males perform household tasks differently from females; therefore products should be positioned differently for males and females.

A recent study (Hendrix and Qualls 1981) sounds a warning that males may not be doing as much household work as they claim, or perhaps as they think they do. Research which is yet unpublished finds that female performance of household tasks has lessened, not just as a result of increased male assumption of household tasks, but as a result of smaller families, use of time-saving products, and changed standards and values relative to household care principally among working women (Pleck 1981). A number of commercial studies, including those cited above, demonstrate that an increasing number of males shop for groceries, and that the shift from female to male performance of the shopping task frequently results in different brands being purchased (see also Newsweek 1979).

In behavioral terms, these (and other) studies suggest that both role transfer and changes in role performance are taking place. Some female roles are being transferred to men who perform them differently than women have in the past; some females who retain particular roles now perform them differently. The situation presents complexities with which research has not get satisfactorilY dealt.

This paper represents an attempt to establish a conceptual framework which may assist in unravelling some of the complexities in the area of changing household standards and roles. We will first present an overall model. We will then test some of its major components within the limits of an existing data set.


The consumer behavior literature does not yet include specific models of adult role learning or role transfer. Even the investigation of relationships between incumbents in different roles seems to be more prevalent in studies of industrial buyer behavior than in studies of household or family consumer behavior. To the extent that the literature does deal with various aspects of roles, theories from the behavioral sciences are presented, with marketing studies being used to illustrate specific applications (see Zaltman and Wallendorf 1979, Chapter 7 for an example of such an approach).

For a conceptualization of role transfer [We are using the concept of role transfer rather than that of role acquisition because we are not so much interested in the general phenomenon of adding a role to an existing role set but in the more specific situation of a role being passed from one person to another. The original role incumbent could be defined either on the basis of previous role behavior, as in an established marriage which has maintained traditional patterns of role performance, or on the basis of traditional stereotypes, as in the case of a newly-married couple who are establishing non-traditional patterns of role performance.] we therefore turned to the sociological and management literature in organizational behavior. Two models are especially useful, Kahn et al.'s (1964) role episode model and Schuler's (1979) transactional process model. In addition, we gained important insights from a paper by Israel (1966) on role learning. The result is the model Presented in Figure 1.

The model assumes two actors -- a role sender and a recipient, or focal person. In a household role transfer situation the role sender would ordinarily be the wife and the husband would be the focal person. However, to the extent that women are assuming traditionally male tasks such as care of an automobile or financial decision-making, the sender/focal roles could be reversed. For the purpose of our discussion and subsequent data analysis we will assume a female role sender and a male focal person since that would appear to be the most common situation with respect to day-to-day household roles. The model is based on the perceptions and expectations possessed by both sender and recipient. The perceptual set is organized around the two basic dimensions of a role, its technical/ instrumental aspects and its expressive/ideological aspects. The technical aspects relate to "knowledge and skills necessary or desirable to carry out the tasks" and the expressive aspects are composed of "attitudes, values, and ideals related to or being the basis of role-related task performance ( Israel 1966, p. 201).

The expectation set is also two-dimensional. One dimension consists of expectations concerning the manner in which the task will be performed. The other set of expectations, or standards, teals with the actual outcome, or result; of performing the task.

In discussing the application of this model to household tasks, we will concentrate first on women as role senders and then on men as recipients. In the discussion, both the content of the role and the transfer process will be considered. Although there is empirical support for the concepts used in the literature on formal organizations, our application to household task performance represents an extension of these concepts to the area of consumer behavior. The discussion to follow will cover both sender's expectations and receiver's task performance; the subsequent partial test of the model will cover only receiver's task performance since we did not have data with which to test sender's expectations against receiver ' s task performance.




Many women have learned the technical aspects of household roles in childhood. However, time-saving products of all kinds have greatly changed the manner in which household tasks are, or at least can be, performed. Therefore, much of women's early learning about the technical aspects of task performance may have been rendered potentially obsolete. Moreover, the mass media are probably very important in facilitating adult learning about more efficient ways of performing tasks.

The expressive aspects of roles and their related tasks are also learned in early childhood. Most women have been socialized, explicitly or implicitly, to regard the nurturing, supporting aspects of caring for their families as important components of their role and self-definitions. As a result, both task performance and outcome have had emotional connotations in addition to their technical dimension. Societal role definitions (or stereotypes, if you will) reinforce the importance of the expressive aspects of task performance. The mass media also subtly reinforce the expressive aspects of certain specific roles.

The nature of the role being transferred influences the nature of the process. The extent to which the task has been gender-linked, its complexity, and the extent to which it offers an opportunity for expressiveness will all affect the ease and manner in which it is transferred.

The two-dimensional nature of roles suggests the possibility of two types of communication, direct and indirect, in the transfer process. Direct communication via verbal instructions, by example, or by assistance with initial task performance seems more appropriate for the transmission of the technical aspects of a role. Indirect, probably even nonverbal, communication seems more likely to take place with respect to the expressive aspects of a role. For one thing, many people have probably never verbalized expressive role content, even to themselves. In addition, many of the expressive aspects of roles sound rather silly when actually put into words - witness some of the findings of motivational research.

Because the expressive aspects of a role are not easily communicated verbally, there is great opportunity for ambiguity and conflict in their transferal. This is especially true since male perception of the nature of household roles is likely to be quite different from female perception of them. Most adult males have not been trained in childhood to perform these roles. Neither have they been socialized to recognize or to value the expressive aspects of performing these roles. As a result, the male focal person's initial understanding of both the technical and the expressive dimensions of a task are likely to differ from those of the sender. Since communications s both from role senders and the media are more likely to concern technical rather than expressive topics, senders and focal person's perceptions of the technical dimension are more likely to be similar than are their perceptions of the expressive dimension.

Actual task performance will result in either satisfaction or dissatisfaction on two dimensions, the manner in which it was performed and the outcome, on the part of both the focal person and the role sender. Satisfaction of both actors on both dimensions results in a satisfactorily transferred role.

Except for roles with very high expressive content, such as child care, we can assume that the role sender is concerned more with outcome than with performance. However, performance that results in an outcome satisfactory to the role sender would ordinarily result in a desire to see the role filled permanently by the focal person. An unsatisfactory outcome could lead to another attempt to effect a satisfactory transfer or to a reassumption of the role by the sender.

The focal person, since he is now performing the task, should be concerned about both performance and outcome. Satisfactory outcome/unsatisfactory performance could lead either to an attempt to acquire more technical information or to an attempt to shift the role back to the sender. Unsatisfactory outcome/unsatisfactory performance might result in similar strategies with perhaps more emphasis on attempting to shift the role. Unsatisfactory outcome/ satisfactory performance is likely to be a null set.

To further complicate matters, the standards of sender and focal person for judging both performance and outcome may differ. Male attitudes toward performing many of these tasks are likely to be at least somewhat negative and males are therefore likely to find performance onerous. Their dislike of performing these tasks is probably not alleviated by pleasure in their expressive aspects as it has been for women -or at least as women have been taught that it should be. Male outcome standards may or may not be lower with respect to technical aspects, but they may be quite different on expressive aspects.

Although many of the concepts used in our model can be supported by studies in other disciplines. the applications we are making to family behavior need to be empirically tested. The technical/expressive dichotomy appears especially promising in terms of increasing our understanding of consumer behavior.

In the absence of data which allows us to thoroughly test the model, we will state a few of the many hypotheses which could be drawn from it and test them with available data. Even limited support for these hypotheses will indicate that the model is worthy of continued research.


H1: Women will show more agreement with statements which describe the expressive aspects of household task performance than will men.

H1a: Working women will show less agreement than non-working women with the expressive aspects of household task performance.

H2: Men and women who actually perform a particular role will not differ significantly in terms of reported performance of technical aspects of the role.

H3: Men and women who actually perform a particular task will indicate similar levels of knowledge about the technical aspects of performing the task.


The hypotheses just stated will be tested using data on the performance of one important household task, food shopping. The data to be analyzed are taken from the 1980 Needham, Harper & Steers Life Style study. This study uses the Market Facts mail panel. Needham, Harper and Steers describe the data base as follows:

"The Market Facts' mail panel is balanced for geographic region, age,income, and degree of urbanization. The very poor, the very rich, the transient, and minority populations are not well represented in the panel. Needham, Harper and Steers places a further restriction on the sample by requiring all individuals to be married. This latter restriction coupled with the general characteristics of the Market Facts' mail panel, tends to confine the representativeness of the Life Style sample to stable, middle class households... (it) has proven to be an effective barometer of mainstream middle America."

The data to be analyzed are from the AIO section of the questionnaire and consist of statements selected from that section which measure technical performance, expressive performance, and role knowledge. Some readers will undoubtedly wish to quarrel with the assignment of particular items to particular categories. But, as will be obvious from the results we present, the reassignment of questionnaire items will not materially affect the conclusions we have reached. It is possible to make comparisons between male and female respondents using these data because questionnaires were assigned randomly by sex within the sample of households. The statistical technique employed will be analysis of variance using sex of respondent and whether the wife works as independent variables. The analysis is confined to heavy shoppers of both sexes, defined as those who shop for groceries more than three times a week, (n - 195 for men and 441 for women).


Tables 1 through 3 present the findings for each hypothesis. The items are listed according to the magnitude of the absolute difference in the percent of males and females agreeing with each item. The tables show the significance of the F statistic for the main effect of sex and wife's work, the significance of sex and wife's work separately, and the significance of the interaction of sex and the wife's work. The table also shows the proportion of variance in each AIO item explained by the two independent variables and the percentage of both male and female agreement with each statement.



The analysis is based on individual AIO items. Because of the exploratory nature of the research, we did not construct scales for each dimension but chose rather to present item data with the hope that it would provide better insights into further research. The direction of the male/female differences is not consistent, and scales would have masked the avenues for additional research that are suggested by careful examination of the item data.

Table 1 shows results for the expressive dimensions (H1 and H1a). Only two of the five items showed significant differences between men and women and only one showed an interaction with wife's work. Moreover, both the two items which showed significant differences had low R2's, and the differences between the percentage of males and females agreeing with each item is not large. Thus, it is tempting to suggest that hypotheses 1 and la should be rejected.

On the other hand, one item that intuitively might be considered the very best indicator of the expressive dimension in the set, "kitchen is my favorite room, showed both the largest difference between the sexes and an interaction with wife's work. This particular result suggests that items specifically designed to measure expressive aspects of the task might have shown much stronger differences between men and women, and we therefore conclude that H1 and H1a should not be rejected on the basis of this data and that they are worth further testing.

The results for the technical performance (H2) items are shown in Table 2. Six of the ten items show statistically significant differences between the sexes (only three show differences between working and non-working wives). The R2's are, once again, extremely low. However, the differences between the male and female agreement scores are rather large on most of the items. They also are rather consistent, with females shoving more agreement with the attitudes or behaviors than males on nine of the ten items. This strongly suggests that men and women to perform the cooking and shopping tasks differently. Therefore R2, which posits no difference, is not supported by these data. Moreover, the data suggest the character of the differences. Men appear to be more empirical in their performance, and at the same time spent less effort.



Knowledge related items (!13 ) are presented in Sable 3. Four of 12 items showed statistically significant differences between men and women; only one showed a significant effect of wife's working, and the R2's are once again extremely low. This lends support to H3 which posits no difference in task-related knowledge between males and females who actually perform the task.

However, when the percentage differences be tween male and female agreement are considered, this conclusion becomes questionable. The differences, overall. are smaller than on the technical performance items, but their nature is intriguing. If one looks at the differences between male and female beliefs on the four items that relate to type of brands (national, store, generic), inconsistency is immediately apparent. Men are more likely to believe that national brands are better than store brands or generics but that generics are better than store brands. The intransitivity in the expected ranking of national, then store, then generic brands leads us to ask whether men or women are more likely to have perceptions of the quality of types of brands that differ from the expected ordering.



There are also differences between males and females that relate to information seeking and dissemination. Women seem to want and use more information from advertising, but are less likely than men to believe the test results that are used in advertising. Information from personal sources appears to be less important to both but males report a slight tendency to seek information more often while females report an equally slight tendency to give it more often. An explanation for these findings that would be consistent with the data is simply that women are more knowledgeable and experienced than men in dealing with cooking and with information about foods and cooking.

These differences are neither clear nor consistent enough to be used as a basis for managerial strategy at present. However, they do suggest that both the content of male/female beliefs about brands and the nature of male versus female information seeking and use should be further investigated. They also suggest that it would be inappropriate to accept H3 on the basis of these data.


We have tested part of a comprehensive model of role transfer using secondary data which were imperfect and confined to one task. We believe the results presented, while far from conclusive, suggest the model may be valid and therefore worth further testing. The most clearly expressive item on the expressive list showed the largest differences between men and women. The extent of the differences on the technical performance dimension was unexpected as were both the nature and extent of the differences on the knowledge dimension. The fact that these items taken separately explained so little of the variance in actual task performance underscores the need for more carefully conceived and detailed empirical testing.

Further research should be both intellectually interesting and managerially worthwhile. Interesting research might be pursued in several directions. One direction would be to study more tasks. Food shopping is probably one of the easiest female roles for men to take on and may also have less significant expressive content than, say, child care or even cooking. A study which compared several tasks would be likely to pick up some of these considerations.

A second direction for further research is to look at within as well as between-sex differences, on both knowledge and technical performance and, even more importantly, expressive performance dimensions. As we suggested earlier, many women may not be as ego-involved with household tasks as they have been in the past. We attempted to control for this possibility by comparing working versus nonworking wives, but this comparison is very rough and there is no reason to suspect a one-to-one correspondence between outside work and lack of involvement in household tasks. Whether men are ego-involved in these tasks at all is still an open question.

In addition, this study did not consider the outcomes dimension. It would be worthwhile to address future studies to this dimension for several reasons. One is to look at the relationship between desired outcome levels and the willingness to transfer a role. Another is to determine whether expectations change upon transferring (or accepting) a role.

All of the research just suggested should have managerial as well as intellectual utility. One important use of these findings would be in positioning. It is often suggested that much of the advertising for packaged goods which are used in the performance of household tasks concentrates on subjective benefits, using appeals that are clearly addressed to the task's expressive dimension. Obviously if there are now changes in the expressive nature of tasks, there need to be changes in advertising as well. These changes include more emphasis on objective benefits as well as more recognition of males as task performers.

Similarly, advertising often assumes that best outcomes are really important, e.g., the whitest clothes, the most spotless drinking glasses. To the extent that expectations for outcomes are lowered or changed as roles are transferred, then advertising must be changed. In some circumstances, product changes are indicated. For example, where there is a trade-off between ease of use and "superior" outcome, the trade-off should now be in the direction of ease of use.

This line of reasoning also suggests increased emphasis on point of purchase information which conveys technical performance information, especially ease of use. Hales who lack both experience in task performance and knowledge about the products used in task performing and who are uninterested in traditional advertising messages may be much more prone to rely on point of Purchase information.


The absence in the marketing literature of an organizing framework for research into the transfer of household roles and tasks has lead us to develop this model. Even though our partial test was not definitive, we hope to encourage a more conceptual approach to research in this area. The information managers use as they make decisions on the attribute/benefit composition and the positioning of household products can be greatly improved by an approach grounded in tested concepts. We offer this paper as a step in that direction.


Benton and Bowles (1980), "Men's Changing Role in the Family of the 80's," An American Consensus Report.

Cunningham and Walsh (April and October, 1980), "Husbands as Homemakers I and II."

Hendrix, Phillip E., Kinnear, Thomas C., and Taylor, James R. (1979), "The Allocation of Time by Consumers: A Proposed Model and Empirical Test," Working Paper No. 188, The University of Michigan.

Hendrix, Phillip E. and Qualls, William J. (1981, forthcoming), "Assessing the Validity of Subjective Measures of Household Task Responsibility with Time-Budget Data."

Israel, Joachim (1966), "Problems of Role-Learning," in Joseph Berger et al., (eds.), Sociological Theory in Progress, Boston: Houghton Mifflin Company.

Kahn, R., Wolfe, D., Quinn, R., Snoek, J. D., and Rosenthal, R. (1964), Organizational Stress: Studies in Role Conflict and Ambiguity, new York: John Wiley and Sons.

Newsweek (1979), "The Role Men Play in Brand Selection of Food Products," A Research Report.

Pleck, Joseph (1981), personal communication with the author.

Roberts, Mary Lou and Wortzel, Lawrence H. (1980), "Husbands Who Make Dinner: A Test of Competing Theories of Marital Role Allocation," in Jerry C. Olsen, (ed.), Advances in Consumer Behavior, Vol. VIIS pp. 669-671.

Schuler, Randall S. (1979), "A Role Perception Transactional Process Model for Organizational Communication-Outcome Relationships," Organizational Behavior and Human Performance, 23, pp. 268-291.

Time (undated), "The Changing World of Men and Women in Brand Selection."

Zaltman, Gerald, and Wallendorf, Melanie (1979), Consumer Behavior Basic Findings and Management Implications, New York: John Wiley and Sons.