Family Type Effects on Household Members’ Decision Making

ABSTRACT - This study investigates the influence of the long-term properties of a family, the so-called 'family-structure’, on family buying decision processes. The results of an empirical survey involving 74 Dutch families showed that on average the children’s influence was relatively high in the later stages of the decision process at the expense of the wife’s influence. However, children in modern families had more influence than those in traditional families in the problem recognition stage. Cohesive families relatively often evaluated alternatives jointly and took into account another’s desires. Non-cohesive families relatively often had conflicts and formed coalitions to solve conflicts.

Citation:

FrTderique Holdert and Gerrit Antonides (1997) ,"Family Type Effects on Household Members’ Decision Making", in NA - Advances in Consumer Research Volume 24, eds. Merrie Brucks and Deborah J. MacInnis, Provo, UT : Association for Consumer Research, Pages: 48-54.

Advances in Consumer Research Volume 24, 1997      Pages 48-54

FAMILY TYPE EFFECTS ON HOUSEHOLD MEMBERS’ DECISION MAKING

FrTderique Holdert, Erasmus University, the Netherlands

Gerrit Antonides, Erasmus University, the Netherlands

ABSTRACT -

This study investigates the influence of the long-term properties of a family, the so-called 'family-structure’, on family buying decision processes. The results of an empirical survey involving 74 Dutch families showed that on average the children’s influence was relatively high in the later stages of the decision process at the expense of the wife’s influence. However, children in modern families had more influence than those in traditional families in the problem recognition stage. Cohesive families relatively often evaluated alternatives jointly and took into account another’s desires. Non-cohesive families relatively often had conflicts and formed coalitions to solve conflicts.

INTRODUCTION

Research on family decision making has developed in several ways. To assess the role of the family members in the decision process, many studies have focused on the (relative) influence of the family members and how the influence varied with the type of decision, decision stage or subdecision (Davis, 1970; Davisand Rigaux, 1974; Wilkes, 1975; Davis, 1976). Furthermore, research has dealt with the factors determining the relative influence of family members, such as the importance of the outcome of the decision, the knowledge of the product and the fact whether the product is for one’s own use (Corfman and Lehmann, 1987; Mittal and Lee, 1989; Foxman, Tansuhaj and Ekstrom, 1989b).

Another development is the acknowledgement of children’s influence in family decision making, whereas in earlier research a dyadic representation of the family was studied (including spouses only). At first, only the husband and the wife were described as the family (Davis, 1970; Davis, 1976) and thus either husband or wife or both were interviewed. With the growing influence of children and adolescents in family decision making, they were gradually taken into account in assessing the different roles and influences within families (Szybillo, Sosanie and Tenenbein, 1977; Belch, Belch and Ceresino, 1985; Foxman, Tansuhaj and Ekstrom, 1989a, 1989b; Mangleburg, 1990; Beatty and Talpade, 1994).

Several factors influencing the family decision making process have been proposed, such as power, parental style and communication patterns. Moreover, the influence of a person on the decision process depends on the importance to that person of the good involved (personal relevance; Mangleburg, 1990; Corfman and Lehmann, 1987; Beatty and Talpade, 1994). In general, research did not go further than describing factors and describing the situation in which family decision making took place. However, as Johnson, McPhail and Yau (1994) state, it is not enough for marketeers only to know how much influence children have on family decisions. Marketeers should understand the behavioral interaction between mother, father and children. Only when these interactions are known, are attempts at influencing the decision making process realistic. In our study, we use information from both children and their parents to investigate their influence in various stages of family buying processes.

The structure of the family, its long-term properties, determines the behavioral interactions (the dynamics) between family members in the decision process. In our study, two structural dimensions, power and cohesion are used to describe the type of family. The effects of these dimensions on four stages of family purchase processes are investigated. Four different purchases are studied: holidays, adult and child clothing and sandwich filling. We assume that if behavioral interactions differ across family types, the buying process also differs across these family types. Indeed, we find that modern and traditional families differ with respect to the influence of family members in the problem recognition stage. Also, we find differences regarding strategies used in decision making.

Next, the two dimensions determining the family type will be considered and several hypotheses regarding the family purchase process will be formulated. The methodology of the study will then be dealt with and the results of an empirical survey will be shown. Finally, conclusions for the study of consumer behavior and marketing will be drawn.

TWO STRUCTURAL DIMENSIONS

The structure of the family consists of the long-term properties of the relationship between family members. Family dynamics consist of the interactions between family members. [Kirchler (1989) states that the current structure forms the starting point of the interaction; it influences the way in which the interaction proceeds. On the other hand, the current structure is the result of interactions in the past. So, in order to avoid drawing tautological conclusions it is important to restrict consequent interactions to one specific area of interaction, namely the family buying decision process.] Several authors have attempted to describe families by means of specific variables. For instance, Olson and McCubbin (1989) distinguished cohesion, the ability to adapt and communication style. Fitzpatrick (1984) described autonomy versus interdependance, conflict-involvement versus conflict-avoidance and convention versus tradition. Kirchler (1989) argued that in order to comprise all these factors, two underlying structural dimensions can be distinguished: power and cohesion.

Power

In the literature, there is no single definition of power. However, there is agreement about the difference between potential power and power actually exerted, called influence (Corfman and Lehmann, 1987). Potential power is the ability of a person to change the attitudes, opinions or behavior of other people. Influence is the consequence of the active or passive exercise of power. In a relationship, the division of power has been described as traditional versus modern, considered to be a continuum. The distribution of power in the relationship affects the way decisions are made.

Traditional distribution of power. According to Ferber (1973), the responsi bility of the husband in a traditional relationship is earning the money whereas the wife is responsible for housekeeping and child care. Davis (1976) claims the existence of large authority differences in traditional relationships. This may frequently take the form of a hierarchical structure. According to Hagenaars and Wunderink-Van Veen (1990) a strictly hierarchical family has a patriarchal structure where the husband and father is considered the head of the family. Kirchler (1989) states that many decisions in a traditional relationship are taken autonomously by one of the spouses. Also, in traditional families coalition formation is likely, as the 'weak’ partner will try to find support from other family members.

TABLE 1

CLASSIFICATION OF FAMILY TYPES

Modern distribution of power. Modern relationships are characterized by a high degree of joint participation in carrying out tasks and taking decisions (Davis, 1976). Hagenaars and Wunderink-Van Veen (1990) state that husband and wife have equal influence in a modern relationship. Also, the power distance between parents and children is shorter in modern than in traditional families. Kirchler (1989) states that in egalitarian relationships, spouses decide much less by role-segregation than in traditional relationships. However, both spouses wish to fulfill their individual desires. As a consequence, many decisions are made together but conflicts arise due to different opinions. Summarizing, these descriptions lead to the classification shown in table 1.

The family type is a structural dimension and has been defined generally. As a hypothesis regarding the family purchase process, we expect the influence on decision making to show a relatively large variation in traditional family types and a relatively small variation in modern family types.

Cohesion

The second underlying structural dimension is cohesion, which is considered a continuous dimension. Cohesion is indicated by the degree of harmony in a family, the degree of interest in each other and the coalition dynamics. In general, cohesion includes the (degree of) emotional bonds between family members. Research has shown that a higher degree of cohesion is associated with a more harmonious family life and less egoistic decision making by the family members (Kirchler, 1989). See table 1.

As a hypothesis relating the structural dimension to family purchases, cohesive families are expected to evaluate alternatives jointly, to take into account another’s desires and to be involved in conflicts less often than non-cohesive families. Coalition formation is expected less often in cohesive families than in non-cohesive families, since the former are more likely to form the 'grand coalition’ involving all family members. (See Murnighan, 1978 for an overview of coalition formation models.) Regarding the type of conflict resolution strategies used in different family types, we have no a priori expectations.

METHODOLOGY

A survey was carried out among 103 families, June 8-29, 1995. In each family, one child and its parents were questioned, to obtain a clear picture o the family as a whole. Children aged 8-12 years were selected since they are able to think conceptually and they can finish concrete tasks by themselves (Ginsberg and Opper, 1969). Beyond the age of 12 years, children are in their puberty, leaving primary school, etcetera. The children were approached at three different schools. They completed the questionnaires in each of their classes in order to avoid problems with the younger children. Each child was asked to take home two parents’ questionnaires with an accompanying letter including a request to complete them separately and independently. No other children or family members were questioned.

The parents’ questionnaires were divided in two parts. In the first part, the family buying decision process was dealt with stage by stage: the problem recognition stage, information search, evaluation of alternatives and decision/purchase. In the second part, questions were asked to determine the family type, following methods used by, for instance, Moschis and Moore (1979) and Carlson, Grossbart and Tripp (1990). To categorize families into these types, statements referring to both dimensions were included. Agreement with the statements was assessed by means of 4-point Likert scales. For example, division of power was measured by questions such as: "Within our family, the parents have the final word, whereas the children have no say" and "I believe that my child should treat me with respect." Cohesion was measured by questions such as: "The relationships within our family can be described as very strong" and "I see my family functioning as a team rather than as a group of independent individuals." Then, following Moschis and Moore (1979) the scales were summed and families were classified as 'high’ or 'low’ on each dimension by splitting up at the median.

Davis (1976) concluded that the influence of the family members varies across product categories. We selected four different goods to represent four different buying processes: family vacation (specialty good), adult clothes and children’s clothes (shopping goods) and sandwich filling (convenience good). The goods selected differ with respect to the length of the decision process and their importance to the family. The goods have been used in earlier research and typically are family goods, frequently requiring joint decision making, not particularly wife-dominated.

TABLE 2

INFLUENCES OF FAMILY MEMBERS ON THE DECISION PROCESS REGARDING FOUR COMMODITIES, AVERAGED OVER FOUR DECISION STAGES

Three versions of the questionnaire were constructed. Each version included questions about the decision process regarding three of the four goods, for example: "Are you allowed to take part in a conversation about the shop to buy your clothes?" (children’s questionnaire) and "Please indicate how much influence different family members have on the joint evaluation of child clothing" (parents’ questionnaires). The children’s influence on the buying process was measured by asking how much influence they thought they had themselves. The parents were asked to report both their own influence and the perceived influence of their spouses and the children concerned.

RESULTS

The response rate to the questionnaire was 74.7%, involving 77 families. Due to partial non-response and incomplete families, observations included 71 males, 74 females and 77 children. On average, the husband’s age was 41 years, the wife’s age was 39 years and the child’s age was 10 years. The average number of children in the families was 2. Regarding schooling, 47.8% of the husbands and 26.4% of the wives had completed higher vocational education or university.

The unit of observation was the family and responses given by the members were considered as family variables. The type of family with respect to tradition and cohesion was classified on the basis of the average score of the spouses on the questions concerned. For incomplete families, the score was based on the score f one parent only. The correlations between spouses’ scores were 0.62 and 0.45 for tradition and cohesion, respectively.

The influence of the family members on the decision process regarding the four commodities was measured by averaging the influence reported by the focal individual, and the influence of the focal individual as perceived by the second and third members of the family, each measured by means of 4-point scales. In the case of missing information, the average was computed using information from the remaining individuals. The correlations of measures from the spouses were generally in the 0.20 - 0.60 range. Those between spouses’ and children’s opinions were generally in the 0.00 - 0.20 range. Table 2 shows the influence of family members, aggregated over the four stages of the decision process.

Since each family answered questions concerning three issues only, a full factorial analysis would have resulted in substantial missing information. For this reason, four MANOVA’s with repeated measures designs were run to test the significance of the family member effects for each issue separately. In each MANOVA, the families were considered as the observation units. Children clearly exerted the least influence in decision making and the wife had the most influence for all four issues. It appeared that children’s influence was the strongest for child clothing, followed by sandwich filling. Apparently, the wife’s influence on child clothing was much stronger than the child’s own influence. The husband’s influence on holiday decision making was relatively high as compared with his influence on other decisions. Negotiation theory predicts more influence if the importance of the issue is high (e.g. Belch, Belch and Ceresino, 1985; Foxman, Tansuhaj and Ekstrom, 1989; Corfman and Lehmann, 1987; Beatty and Talpade, 1994). In our survey, we asked each family member how important the issues were to themselves by means of 4-point Likert-type scales. Although the average influences varied considerably, we found little variation in average importance of the issues:

- for holidays, the average importances reported by the family members individually were in the 3.20-3.22 range;

- for sandwich filling, the average importances were in the 2.63-2.92 range;

- for adult clothing, the average importance range was 2.49-2.55;

- for child clothing, the range was 3.06-3.27.

It appeared that the variation in influence was virtually unrelated to the variation in issue importance, contrary to the prediction from negotiation theory (Corfman and Lehmann, 1987; Beatty and Talpade, 1994).

The influence of the family members may vary across the different stages of the decision making process. Table 3 shows the average influence of the family members across the four issues.

A full factorial MANOVA with a repeated measures design was run to test the significance of the effects. Both the main effects and the interaction effect were statistically significant. In general, the husband had less influence in the problem recognition stage than in the other stages. The wife had the most influence in general and particularly in the first two stages of the decision making process. Children had more influence in the stages of alternative evaluation, choice and purchase than in problem recognition and information search. These results also hold for the relative influence of the family members, i.e. as percentages of total influence perceived. In general, the alternative evaluation stage yielded the highest overall perceived influence, possibly due to its impact on the final outcome of decision making.

TABLE 3

INFLUENCES OF FAMILY MEMBERS ON THE STAGES OF THE DECISION PROCESS, AVERAGED OVER FOUR COMMODITIES

TABLE 4

INFLUENCES OF MEMBERS OF TRADITIONAL AND MODERN FAMILIES IN THE FIRST STAGE OF DECISION MAKING, AGGREGATED OVER ISSUES

Tables 2 and 3 showedthe influence of family members regardless of family type. To study the effect of family type on decision making, traditional and modern families were distinguished such that modern families scored above the median of the distribution (24, summed over 9 items) and the remaining families were considered traditional. The analyses above were repeated with the modern versus traditional distinction added as a between-subjects factor in the MANOVA’s. [A MANOVA including both the traditional versus modern and the cohesion dimensions resulted in non-significant cohesion effects, which are not reported here.] Family type was not found to be a significant factor in the influence of family members on the aggregated decision process per issue (see table 4). With the influence aggregated over issues, the interaction of family type, family member and decision stage yielded an F-ratio of 1.94, df(6,63) which was significant at the 10% level. The interaction effect was evident mainly in the first decision stage (problem recognition). The influence of family members in the problem recognition stage (aggregated over issues) is shown in table 4. In addition, we computed correlations between the influence of each member and the 'score’ on the traditional versus the modern scale, which are reported in table 4. It appeared that, on average, the husband’s and the child’s influence in problem recognition were stronger in modern than in traditional families, whereas the influence of the wife was less strong. So, in modern families the power was distributed more equally than in traditional families in the problem recognition stage. In the other stages, no significant differences were found. Our hypothesis regarding the effect of family type on decision making has only partially been confirmed.

Next we considered the strategies of decision making used in weakly and strongly cohesive families. Cohesion was correlated with the following variables: individual versus joint evaluation of alternatives, taking another’s desires into account to a large or small extent, whether or not conflicts arise sometimes, how often coalitions are formed (i.e. including the opinion of a third member of the family in the conflict) and the use of other problem solving strategies. These variables were constructed at the family level using the information from both spouses as follows: for 2-point scales (whether or not conflicts arise), disagreement between spouses was scored in between the values of the two points; [The 2-point scale served as a selection device for further questions regarding conflicts.] for 3-point scales (the other variables, except coalition formation), extreme disagreement was scored at the middle of the scale and disgreement by one point was scored in between the values of the two points concerned. The 4-point coalition formation scale was simply averaged over the spouses. For each consumption issue, Spearman rank correlations between cohesion and the type of strategy are shown in table 5.

TABLE 5

RANK CORRELATIONS BETWEEN COHESION AND STRATEGIES USED IN DECISION MAKING, REGARDING FOUR ISSUES

TABLE 6

CONFLICT RESOLUTION TACTICS USED BY WEAKLY AND STRONGLY COHESIVE FAMILIES

It appeared that strongly cohesive families evaluated alternatives jointly more often than weakly cohesive families, except for adult clothing which is after all a personal consumption item. Furthermore, it appeared that strongly cohesive families took into account another’s desires relatively often, had fewer conflicts and less often formed coalitions regarding sandwich filling. This is in agreement with our expectations.

Finally, spouses were asked which tactics were used in case of conflict: discussion and gathering more information, persuasion, negotiating (trading off issues) and other tactics (see table 6). It appeared that discussion and gathering information was used most, followed by negotiation. Other techniques were used relatively often in strongly cohesive families. The latter result was not statistically significant, however.

CONCLUSIONS

The wife’s influence on household decision making appeared to be considerably stronger than the husband’s and the latter’s influence appeared to be stronger than the children’s. This could hardly be explained by product type since typical family products were selected, the choice of whch is not necessarily wife-dominated. In the problem recognition stage, the wife’s influence in modern-type families turned out to be slightly less than her influence in traditional types, although it remained considerably stronger than the husband’s. This suggests that the modernization of the Dutch family is only beginning.

The influence of children appeared to be higher in the evaluation and decision stages than in problem recognition and information search, contrary to findings by Belch, Belch and Ceresino (1985). We cannot rule out the possibility that the type of product has mediated this effect. In general, for low interest products and in democratic families, we assume that parents would allow their children to influence the decision. Although advertising may influence children’s knowledge and preferences regarding alternatives in unrestricted choice (Gorn and Goldberg, 1982; Goldberg, Gorn and Gibson, 1978), its effect on family decision making seems limited, given the children’s low influence in the first stages of decision making.

Cohesion was associated with decision making strategies in several instances. Cohesive families relatively often evaluated alternatives jointly, frequently took into consideration another’s desires and ran into conflicts less often. Problem solving tactics were used relatively often in our sample, confirming Johnson’s suggestion (1995) that this strategy will be used frequently in triadic relationships. This finding is in contrast with Belch, Belch and Schiglimpaglia (1980) who found bargaining and persuasion strategies in triads relatively often. A future research question may include the agreement of family members’ perceived influences in relation to cohesion.

The children’s influence on child clothing was stronger than on the other issues, suggesting some association between issue importance and influence. Importance of clothing might be partly due to its symbolic value (Solomon, 1983). However, the low variation in issue importance was inconsistent with the high variation in influence of the members on the decision process. This may be due to a discrepancy between individually perceived influence and aggregate influence as indicated by the average perceptions of the family members.

Since we have considered family type and the influence of family members as family characteristics, the unit of observation was the family and the variables were aggregates constructed from individual information. Although our research was limited by the number of families in the sample, we are convinced that our approach is to be preferred because the variables are less affected by subjective distortions than by using the individual as the unit of observation.

Although children’s information processing may differ across different ages (Roedder-John and Whitney, 1986; Gregan-Paxton and Roedder-John, 1995), parental style may be roughly constant and children’s relative influence in family decision making may not change much across different family types. The investigation of different age groups should be left for future research, however. Future research may also be aimed at generalizing our results with respect to different age classes, family composition (e.g. single parent families, small and large families), children’s gender, different products and cultures.

Because of the different roles of the family members in the decision making process, marketeers should not consider the family as a whole. Rather, they should make use of the role specializations within the family and of the knowledge that the structure and the dynamics of the family are determining factors in the family buying decision process.

REFERENCES

Beatty, Sharon E. and Salil Talpade (1994), "Adolescent Influence in Family Decision Making: A Replication with Extension," Journal of Consumer Research, 21 (September), 332-341.

Belch, Michael A., George E. Belch and D. Schiglimpaglia (1985), "Conflict in Family Decision Making: An Exploratory Investigation," Advances in Consumer Research 7, 477-478.

Belch, George E., Michael A. Belch and Gayle Ceresino (1985), "Parental and Teenage Child Influences in Family Decision Making," Journal of Business Research, 13, 163-176.

Carlson, Les, Sanford Grossbart and Carolyn Tripp (1990), "An Investigation of Mothers’ Communication Orientations and Patterns," Advances in Consumer Research, 17, 804-812.

Corfman Kim P. and Donald R. Lehmann (1987), "Models of Cooperative Group Decision-Making and Relative Influence: An Experimental Investigation of Family Purchase Decisions," Journal of Consumer Research, 13 (June), 1-13.

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

Davis, Harry L. (1976), "Decision Making within the Household," Journal of Consumer Research, 2 (March), 241-260.

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

Ferber, Robert (1973), "Family Decision Making and Economic Behavior," in: Family Economic Behavior, Problems and Prospects, ed. E.B. Sheldon, Philadelphia: J.B. Lipindt Co.

Fitzpatrick, Mary A. (1984), "A Typological Approach to Marital Interaction: Recent Theory and Research," in: Advances in Experimental Social Psychology, 18, ed. Leonard Berkowitz, New York: Academic Press, 2-47.

Foxman, Ellen R., Patriya S. Tansuhaj and Karin M. Ekstrom (1989a), "Adolescents’ Influence in Family Purchase Decisions: A Socialization Perspective," Journal of Business Research, 18, 159-172.

Foxman, Ellen R., Patriya S. Tansuhaj and Karin M. Ekstrom (1989b), "Family Members’ Perceptions of Adolescents’ Influence in Family Decision Making," Journal of Consumer Research, 15 (March), 482-491.

Ginsburg, Herbert and Sylvia Opper (1969), Piaget’s Theory of Intellectual Development, Englewood Cliffs, NJ: Prentice Hall.

Goldberg, Marvin E., Gerald J. Gorn and Wendy Gibson (1978), "TV Messages for Snack and Breakfast Foods: Do They Influence Children’s Preferences?" Journal of Consumer Research, 5 (September), 73-81.

Gorn, Gerald J. And Marvin E. Goldberg (1982), "Behavioral Evidence of the Effects of Televised Food Messages on Children," Journal of Consumer Research, 9 (September), 200-205.

Gregan-Paxton, Jennifer and Deborah Roedder John (1995), "Are Young Children Adaptive Decision Makers? A Study of Age Differences in Information Search Behavior," Journal of Consumer Research, 21 (March), 567-580.

Hagenaars, Aldi J.M. and Sophia R. Wunderink-Van Veen (1990), Soo Gewonne Soo Verteert. Economics of the Household Sector, (in Dutch) Leyden: Stenfert Kroese.

Johnson, Melissa, Janelle McPhail and Oliver H.M. Yau (1994), "Conflict in Family Purchase Decision Making: A Proposal for an Investigation of the Factors Influencing the Choice of Conflict Resolution Strategies by Children," Asia Pacific Advances in Consumer Research, 1, 229-236.

Johnson, Melissa (1995), "The Impact of Product and Situational Factors on the Choice of Conflict Resolution Strategies by Children in Family Purchase Decision Making," European Advances in Consumer Research, 2, 61-68.

Kirchler, Erich (1989), Buying Decisions in the Private Household, (in German) G÷ttingen: Hubert & Co.

Mangleburg, Tamara F. (199), "Children’s Influence in Purchase Decisions: A Review and Critique," Advances in Consumer Research, 17, 813-825.

Mittal, Banwary and Myung-Soo Lee (1989), "A Causal Model of Consumer Involvement," Journal of Economic Psychology, 10, 363-389.

Moschis, George P. and Roy L. Moore (1979), "Family Communication and Consumer Socialization," Advances in Consumer Research, 6, 359-363.

Murnighan, J. Keith (1978), "Models of Coalition Behavior: Game Theoretic, Social Psychological and Political Perspectives," Psychological Bulletin 85(5), 1130-1153.

Olson, David H., Hamilton McCubbin, Howard L. Barnes, Andrea S. Larsen, Marla J. Muxen and Marc A. Wilson (1989), Families. What Makes Them Work?, Beverly Hill, CA: Sage.

Roedder John, Deborah and John C. Whitney, Jr. (1986), "The Development of Consumer Knowledge in Children: A Cognitive Structure Approach," Journal of Consumer Research, 12 (March), 406-417.

Solomon, Michael R. (1983), "The Role of Products as Social Stimuli: A Symbolic Interactionism Perspective," Journal of Consumer Research, 10 (December), 319-329.

Szybillo, George J., Arlene K. Sosanie and Aaron Tenenbein (1977), "Should Children Be Seen But Not Heard?" Journal of Advertising Research, 17 (6) (December), 7-13.

Wilkes, Robert E. (1975), "Husband-Wife Influence in Purchase Decisions - A Confirmation and Extension," Journal of Marketing Research, 12 (May), 224- 227.

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