Advances in Consumer Research Volume 20, 1993 Pages 469-474
FEMALE-HEADED SINGLE PARENT FAMILIES: AN EXPLORATORY STUDY OF CHILDREN'S INFLUENCE IN FAMILY DECISION MAKING
Roshan "Bob" D. Ahuja, Xavier University
Kandi M. Stinson, Xavier University
This study examines the relationships among selected characteristics of female-headed single parent families, and the influence the children have in the family decision making process. The characteristics of interest are, the mother's age, education, income, sex role orientation, employment status, and the number of years since the mother's separation, divorce, or widowhood, the household size, the age and sex of the oldest child. The results indicate that children's influence in this family type varies according to demographic characteristics of the family, according to the mother's sex role orientation, according to the type of product investigated, and according to the stage in the decision making process.
Consumer behavior researchers have been encouraged to study the family, especially with respect to decision making (Sheth 1982; Davis 1976). Too much consumer behavior theory and research has been focused on the individual consumer and too little attention has been given to the decision making processes involving group behavior, such as those processes of the household (Sheth 1982). Davis (1976) proposed that a theory of household decision making will not emerge by concentrating on decision outcomes, such as who decided or who won. Rather, theoretical progress will be made when more is known about the processes that families use to make decisions.
The most comprehensive existing theoretical marketing model on family decision making (Sheth 1974) assumes the children are growing up in an intact or two parent family structure. The Engel, Kollat, and Blackwell (Engel, Blackwell, and Miniard 1986) model of consumer behavior is oriented around the "normal" (page 270) family structure, i.e., a married couple with children. Ekstrom, Tansuhaj, and Foxman (1987) proposed that a child's influence varies according to family structure. Much empirical information already exists in the marketing literature regarding the child's level of influence in married family households (Atkin 1978; Belch, Belch, and Ceresino 1985; Berey and Pollay 1968; Foxman, Tansuhaj, and Ekstrom 1989; Jenkens 1979; Mehrotra and Torges 1977; Nelson 1979; Szybillo, Sosanie, and Tennenbein 1977; Szybillo and Sosanie 1977; Ward, Wackman, and Wartella 1975; Ward and Wackman 1972); however, very little is known in this area about the female-headed single parent family. The two parent family structure, while still considered the "traditional" or the "normal" family structure by most researchers, is not the typical or modal family structure in the United States in the 1990's.
Marketing academicians and researchers must devote greater attention to the decision making process within the fastest growing family type in the United States, the female-headed single parent family. Too much marketing research has concentrated on the two parent family structure, and too little on the single parent family form, especially the female-headed household. Since 88 percent of all single parent families are headed by women (Statistical Abstract 1989 t67), the study of female-headed single parent families is particularly needed. Since the family is the basic purchase and consumption unit, marketers need to continuously study changes in family structure as they relate to changes in market behaviors.
The specific research objective for this exploratory study is to examine the relationships among selected characteristics of single mother's families, and the influence the children have in the family decision making process. The nine characteristics of interest are, 1) the age of the mother, 2) the education of the mother, 3) the mother's income, 4) the mother's sex role orientation, 5) the mother's employment status, 6) the number of years since the mother's separation, divorce, or widowhood, 7) the household size, 8) the age of the oldest child, 9) and the sex of the oldest child. These objectives were carried out in a national probability sample of female-headed single parent households.
It has been estimated that the 1990 census will show that there are 94 million households in the United States, with 67 million or 71 percent considered family households and 27 million or 29 percent considered nonfamily households (Waldrop and Exter 1990). Of the 67 million family households, approximately 12 percent will be females heading families alone, compared to 5 percent in 1970. Female-headed single parent families represent the fastest growing family type, up 36 percent since 1980 (Waldrop and Exter 1990).
A review of the marketing literature revealed only five studies that, in addition to other issues addressed, also measured the type of family structure and used this measure to investigate the effect that family structure has on selected marketing related variables. With only five studies, it would seem unlikely that similarities exist in the topics studied, but this is not the case.
Similarities do exist along several dimensions. For instance, four studies measured parental perception of the child's influence in the family decision making process (Darley and Lim 1986; Taylor, Moore, and Glynn 1986; Taylor, Glynn, and Taylor 1985; Kourilsky and Murray 1981). One dealt with grocery shopping behaviors specifically (Sinkula 1984) while two had measures for the child's grocery shopping autonomy (Taylor, Moore, and Glynn 1986; Taylor, Glynn, and Taylor 1985). Two studies considered child age as relevant (Darley and Lim 1986; Taylor, Moore, Glynn 1986) while two considered the age of the adult (Taylor, Moore, and Glynn 1986; Taylor, Glynn, and Taylor 1985).
Darley and Lim (1986) investigated parental perception of the child's influence for leisure-time activities and concluded that in the timing of aspects in the decision process, single parents perceive greater child influence. Taylor, Moore, and Glynn (1986) found children from single parent families are much more likely to purchase food products on their own and influence brand choices more than children from the other family structures. Taylor, Glynn, and Taylor (1985) found children in single parent homes had the greatest influence when their parents were younger, had higher income levels or higher educational levels.
Kourilsky and Murray (1981) studied the use of an economic reasoning model in family budgetary decision making and concluded that children in a single parent family may be treated more like adults, may be more likely to be consulted about expenditures, and may be better informed about the limitations of family resources. Sinkula (1984) included only single parents in his population and attempted to differentiate between female and male headed single parent families using four life style constructs. He found that, compared to male single parents, female single parents are more organized and use coupons more. He also found there is an inverse relationship between usage of frozen foods and food shopping preplanning efforts in both types of single parent families.
This exploratory research starts with the proposition that child influence in mother-only single parent households is not homogeneous within this family type. Rather, differences in child influence among female-headed single parent families may be found relative to differences in the nine variables under investigation.
It is hypothesized that child influence in the decision making process increases in an inverse relation to the mother's age, and in direct relation to the mother's education and income (Taylor, Glynn, and Taylor 1985; Taylor, Glynn, and Marlow 1984).
According to Buss and Schaninger (1984), a woman's sex role orientation affects her household task allocation behavior, finance handling, and influence in the family decision making process. Green and Cunningham (1975) found that, in married families, a wife's sex role orientation affects the family's decision making process and purchasing behavior. It is hypothesized that the more liberal a single mother's sex role orientation, the more influence the children will have in family decision making.
Colletta (1983) found divorced mothers working full time lack the option of leaving their children home, and often take them along on shopping trips. These mothers reported shopping took more time, caused more tension, and often involved buying the children things the mother felt she could not afford. It is hypothesized that children will have more influence in single parent families in which the mother is employed compared to the households in which the mother does not work.
In single parent families formed by divorce, the mothers often lose power. Taylor, Glynn, and Taylor (1984) point out that the influence of children expands with the passage of time in the single parent family. Therefore, it is hypothesized that the greater the number of years since the mother's separation, divorce, or widowhood, the greater the children's influence.
It is hypothesized that older children in the mother-only household will have more influence in family decision making than younger children (Burden 1986; Devall, Stoneman, and Brody 1986; Dornbusch et al. 1985; Taylor, Glynn, and Marlow 1984; Weiss 1979).
Peters (1985) found that differences existed in the way single mothers assign household tasks to their male and female children. She reported that single mothers tend to overwork their daughters and underwork their sons. Devall, Stoneman, and Brody (1986) also found gender effects regarding household task allocation in mother-only households. Therefore, it is hypothesized that girls will have more influence than boys.
While measures for the number of children in the household have largely been ignored in marketing studies, there is support in the sociology literature that the number of children affects the process and outcomes of a family's division of labor decisions regarding household chores (Brody 1986; Gonga 1982; McLanahan 1983; Weiss 1979). It is hypothesized that the larger the household size the greater the influence of the children.
Grocery Products Chosen
Grocery shopping is a task every household must perform on a regular basis and children have been shown to exert discernible levels of influence for these types of products. Eleven products were chosen to represent the grocery product domain 1) breakfast cereal, 2) snack foods, 3) candy, 4) soft drinks, 5) hot dogs, 6) luncheon meats, 7) cheese, 8) soups, 9) laundry detergent, 10) housecleaning products, and 11) children's personal grooming aids.
All of theses products have been used in previous marketing research on parent-child interactions in family decision making. For example breakfast cereal was used by Ward and Wackman (1972), Belch, Belch, Ceresino (1985), Taylor, Moore, and Glynn (1986), Berey and Pollay (1968), Atkin (1978); snack foods by Ward and Wackman (1972), Adler (1980), Mehrotra and Torges (1977), Ward, Wackman, Wartella (1975), Taylor, Moore, and Glynn (1986); candy by Ward and Wackman (1972) Ward, Wackman, Wartella (1975), Mehrotra and Torges (1977), Adler (1980); soft drinks by Ward and Wackman (1972), Mehrotra and Torges (1977), Taylor, Moore, and Glynn (1986), Adler (1980); hot dogs by Mehrotra and Torges (1977); luncheon meats by Mehrotra and Torges (1977); cheese by Mehrotra and Torges (1977); soups by Ward, Wackman, Wartella (1975); housecleaning products by Ward and Wackman (1972); laundry detergent by Ward and Wackman (1972); and children's personal grooming aids by Taylor, Moore, and Glynn (1986), Ward and Wackman (1972).
Survey Development and Sampling
An ex post facto research design with cross-sectional survey data was used for data gathering purposes. Revisions were made to the questionnaire following a pretest of the instrument.
The sampling frame consisted of single mother households taken from the national membership list of the Market Facts, Incorporated's Consumer Mail Panel (CMP). The CMP is representative of the geographical divisions in the United States with respect to the characteristics of U.S. households such as age of the panel member, household size, and household income.
A random sample was taken from a population consisting of female-headed single parent families formed by divorce, separation, or widowhood, with at least one child 18 years old or younger living with the mother. A total of 378 surveys were mailed out to mothers heading families alone. A total of 210 surveys were returned, resulting in a 56 percent response rate.
Dependent Variables: Child Influence Measures
The mother's perceptions of her children's level of influence were measured for each of the five decision making stages considered individually, using all eleven products as the dependent variables. The five decision stages were: (1) the initiation stage, (2) the search for information stage, (3) the evaluation of alternatives stage, (4) the final decision stage, and (5) the purchase stage. Child influence was measured through the use of a 100 point constant sum scale. Mothers were asked to allocate 100 points between themselves, their oldest child, and their other children. The children's influence scores were summed into one measure.
Previous research has indicated that child influence varies by product type, therefore a preliminary analysis of the children's influence scores for the eleven products (within each stage) was performed using a principal-component factor analysis with varimax rotation. With the exception of minor differences between stages, an analysis of the factor loadings indicated that certain products clustered into three groups. The specific products in each group paralleled the findings of Mehrotra and Torges (1977).
Table 1 presents the results of the reliability analysis (using Cronbach's Alpha) for each factor within each stage. Factors in Group A included breakfast cereal, hot dogs, luncheon meats, cheese, soups, and the children's personal grooming products; Factors in Group B included laundry detergent and housecleaning products; Factors in Group C contained snack foods, candy, and soft drinks. It was concluded that, given the results of this factor analysis, given the high reliability scores for the three factors, and given the existence of previous research with similar groupings, each of these factors and product groupings represent adequate unidimensionality to be treated as separate dimensions of child influence. Therefore, three separate dependent variables were used for each stage, representing a summed child influence score for each of the three product groupings presented in Table 1.
RELIABILITY SCORES FOR THE THREE FACTORS (CRONBACH'S ALPHA)
Some of the independent variables, e.g., age of the mother, education of the mother, household size, the age of the oldest child, and the sex of the oldest child, were part of the data base provided by Market Facts. Sex of the oldest child was coded as a dummy variable. The education of the mother was provided by Market Facts as categorical data. This variable was recoded to approximate a metric measure. The mid-points for the categorical data were used as an estimate of the mother's educational level.
The remaining variables were collected on the self-report survey instrument completed by the mothers. The mothers provided a dollar figure indicating the amount of income earned by their employment. The mother's were asked if they worked outside the home; this response was recoded as a dummy variable. Mothers also provided the number of years since separation, divorce, or widowhood.
The mother's sex role orientation was measured metrically using Arnott's (1972) feminism autonomy scale. Arnott's scale is a typical scale used in consumer research (Engel and Blackwell 1982). It has been used in studies by Green and Cunningham (1975) and Venkatesh (1980). The mothers were asked to rate their level of agreement on ten items using a Likert-type seven-point scale. The higher the woman's score on all ten items, the more positive her attitude towards women's autonomy. While other scales are available and have been used in marketing research to measure sex role orientation (Gentry and Haley 1984; Qualls 1987; Rosen and Granbois 1983; Spiro 1983; Schaninger, Buss, and Grover 1982), Arnott's has proven to be a reliable, valid, and compact measure in previous empirical research (Venkatesh 1980).
Results of the stepwise regression analyses are presented in Table 2. There are 15 separate regression equations, three for each of the five decision stages. In each stage the eleven products are in one of three product groups, Group A includes breakfast cereal, hot dogs, luncheon meats, cheese, soup, and children's grooming aids, Group B represents laundry detergents and housecleaning products, and Group C contains snack food, candy and soft drinks.
For the Initiation (stage 1), the three product groupings resulted in different levels children's influence. Age of the oldest child positively effects children's influence in group A; while age of the oldest child has a positive influence and the mother's education has a negative influence on the children's influence for products in group B. In Group C, household size has a positive effect on children's influence.
In the Information Search Stage, the age of the oldest child has a positive effect and the mother's education a negative effect on the level of the children's influence for products in Group A. For products in Group B, both the mother's sex role autonomy score and her education have negative effects on children's influence. In Group C, the age of the oldest has a positive effect on children's influence for these products.
In the Alternative Evaluation stage the age of the oldest child has a positive influence and the mother's education a negative influence on children's influence for products in Group A. For products in Group B, the mother's income has a positive effect while her education and sex role autonomy have negative effects on children's influence. There were no variables significant for products in Group C for this stage.
In the Final Decision stage, there were no variables significant for products in Group A. For products in Group B, as the mother's education increases and her sex role autonomy increases, her children's influence decreases. As the age of the oldest child increases, the children's influence increases for those products listed in group C.
In the Purchase stage, as the mother's education increases and her sex role autonomy increases, her children's influence decreases for products in Group A and Group B. For products in Group C, the older the first born, the higher the children's influence, while as the sex role autonomy of the mother increases, the children's influence decreases.
Not all female-headed single parent families are alike with respect to the children's level of influence in decision making. Children's influence in this family type varies according to the type of product investigated, according to the mother's sex role orientation, according to demographic characteristics of the family, and according to the stage in the decision making process. Most of the findings in this study are consistent with previous research on parent-child interactions in decision making; similarities and differences were found and are discussed next.
We found that the children's influence is product-specific, consistent with the findings of other researchers investigating single parents (Darley and Lim 1986; Taylor, Moore, and Glynn 1986; Taylor, Glynn, and Taylor 1985) and consistent with the findings of researchers investigating married families (Belch, Belch, and Ceresino 1985; Mehrotra and Torges 1977; Szybillo, Sosanie, and Tennenbein 1977; Szybillo and Sosanie 1977, Ward, Wackman, and Wartella 1975; Ward and Wackman 1972).
Two demographic predictor variables, the mother's income, and household size, had positive effects on children's influence. In this study the mother's income had an effect only on the children's influence for cleaning products (detergent and housecleaning), and only in the alternative evaluation stage of decision making. Ward and Wackman (1972), studying married families and the only previous parent-child interaction study to consider these two products, did not find statistically significant levels of child influence. Since children in single parent families spend more time on household tasks (such as cleaning), and are held more accountable for the performance of the work (Peters 1985; Weiss 1979), it is possible that the higher income single parent may be financially able to let the child have more influence in the choice of cleaning products used by the family.
REGRESSION OF PRODUCTS AND DECISION STAGES ON 9 PREDICTOR VARIABLES (STATISTICALLY SIGNIFICANT FINDINGS REPORTED)
As mentioned earlier, household size has rarely been included in marketing studies. We found that children's influence increased as household size increased; however, only in the Initiation stage and only for snack products (snacks, candy, and soft drinks). The effect of household size for single parent families may be different compared to its effect on married households. In single parent families the loss of the additional adult (father or mother) means the transfer of influence to the children. This transfer may be more pronounced as the number of children increases. Since single parents tend to shop with their children, it may be that the initial suggestion for these products occurs prior to or during the shopping trip.
Three predictor variables, age of the oldest child, the mother's educational level, and the mother's sex role autonomy resulted in consistent effects on children's influence across the individual stages. As the age of the oldest child increased, children's influence in family decision making increased. As the mother's educational level increased and her sex role autonomy increased, children's influence decreased. While the effect with respect to age of the oldest child was expected and is consistent with previous marketing and sociological studies on the single parent family, what was not expected was the inverse relationship between the mother's education, the mother's sex role autonomy, and the children's influence.
It may be that time constrained single mothers are less able to engage the children directly in discussions about products and purchase decisions. The better educated, more sex role autonomous single mothers, more confident (than the less well educated, less autonomous single mothers) in their ability to make the best decisions for their family, and facing constant time pressures, may seek less direct input from the children. Instead, these mothers are taking the interests and preferences of their children into account, therefore attributing more of the influence to themselves. The result is that the children become less involved in the decision process.
A similar idea was originally proposed by two of the earliest sociologists writing on single parent families, Glasser and Navarre (1965). They noted that, given the single parent's need to accomplish household tasks in an efficient manner in terms of time and energy expended, children may become less involved in some tasks. Colletta (1983) found support for this proposition and reported that there was a trend for children in moderate income single parent families to help themselves the least. The lack of self-help was related to the constant time pressures single mothers were under. Mothers reported it was easier for them to do things for the child rather than expect the child to accomplish the tasks on their own.
A similar conclusion may be reached for this study. Grocery shopping is a household task all family structures must accomplish. Better educated, more autonomous single mothers are placing more of the burden for family decision making (for grocery products) on themselves, and attributing less influence to their children, a tactic that may be easier for the mother in terms of time and energy expended.
(partial list provided, others available upon request)
Arnott, Catherine C. (1972), "Husband's Attitude and Wives' Commitment to Employment," Journal of Marriage and Family 34, No. 4, (November), 673-684.
Colletta, Nancy D. (1983), "Stressful Lives: The Situation of Divorced Mothers and Their Children," Journal of Divorce 6, (Spring), 19-31.
Darley, William K. and Jeen-Su Lim (1986), "Family Decision Making in Leisure-Time Activities: An Exploratory Investigation of The Impact of Locus of Control, Child Age Influence Factor and Parental Type on Perceived Child Influence," Advances in Consumer Research 13, Richard J. Lutz, ed., Provo, UT: Association for Consumer Research, 370-374.
Davis, H. L. (1976), "Decision Making Within the Household," Journal of Consumer Research 2 (March), 241-260.
Foxman, Ellen R., Patriya S. Tansuhaj, and Karin M. Ekstrom (1989), "Family Members' Perceptions of Adolescents' Influence in Family Decision Making," Journal of Consumer Research 15 (March), 482-491.
Glasser, Paul H. and Elizabeth Navarre (1965), "Structural Problems of the One-Parent Family," Journal of Social Issues, 21 (June), 98-109.
Kourilsky, Marilyn and Trudy Murray (1981), "The Use of Economic Reasoning to Increase Satisfaction with Family Decision Making," Journal of Consumer Research 8 (September), 183-188.
Mehrotra, Sunil and Sandra Torges (1977), "Determinants of Children's Influence on Mother's Buying Behavior," Advances in Consumer Research, William Perreault Jr. ed., Atlanta: Association for Consumer Research, 56-60.
Sinkula, James M. (1984), "A Look At Some Shopping Orientations in Single Parent Households," American Marketing Association Educator's Proceedings, Series 50, Russell W. Belk et al., eds., Chicago, IL: American Marketing Association, 22-25.
Statistical Abstract of the United States (1989), U.S. Government Printing Office, Washington, D.C.
Szybillo, George J. and Arlene K. Sosanie (1977), "Family Decision Making: Husband, Wife and Children," Advances in Consumer Research IV, William D. Perrault, ed., Atlanta, GA: Association for Consumer Research, 46-49.
Taylor, Ronald D., Karen Glynn and Nancy Marlowe (1984),"Family Buying Patterns: A Comparison of Dual-Career and Traditional Families," Marketing in a Dynamic Environment, Atlantic Marketing Association, 145-151.
Taylor, Ronald D., Karen Glynn and Jan C. Taylor (1985), "The Influence of Children in the Purchasing Process Used by Families: An Analysis Comparing Children from Three Different Family Structures," Proceedings of Atlantic Marketing Association.
Waldrop, Judith and Thomas Exter (1990), "What the 1990 Census Will Show," American Demographics, 12, Number 1, (January), 20-29.
Ward, Scott and Daniel Wackman (1972), "Children's Purchasing Influence Attempts and Parental Yielding," Journal of Marketing Research 9 (August), 316-319.
Weiss, Robert S. (1979), Going It Along: The Family Life and Social Situation of the Single Parent, New York, NY: Basic Books, Inc.