Exploring the Dimensionality of Children’S Direct Influence Attempts

Laura A. Williams, San Diego State University
Alvin C. Burns, Louisiana State University
ABSTRACT - Understanding children’s purchase influence is an area in need of further research. To address this need, this research develops a scale to measure children’s direct influence attempts based on social power theory. Through semi-structured depth interviews with 20 children and 12 mothers and surveys of 516 children aged 8-11, scales representing seven dimensions of direct influence attempts are developed. These scales include asking nicely, displaying anger, bargaining, showing affection, begging and pleading, just asking, and conning. These seven scales are shown to demonstrate adequate internal consistency, discriminant validity, and construct validity. Future research suggestions and scale applications are also discussed.
[ to cite ]:
Laura A. Williams and Alvin C. Burns (2000) ,"Exploring the Dimensionality of Children’S Direct Influence Attempts", in NA - Advances in Consumer Research Volume 27, eds. Stephen J. Hoch and Robert J. Meyer, Provo, UT : Association for Consumer Research, Pages: 64-71.

Advances in Consumer Research Volume 27, 2000      Pages 64-71

EXPLORING THE DIMENSIONALITY OF CHILDREN’S DIRECT INFLUENCE ATTEMPTS

Laura A. Williams, San Diego State University

Alvin C. Burns, Louisiana State University

ABSTRACT -

Understanding children’s purchase influence is an area in need of further research. To address this need, this research develops a scale to measure children’s direct influence attempts based on social power theory. Through semi-structured depth interviews with 20 children and 12 mothers and surveys of 516 children aged 8-11, scales representing seven dimensions of direct influence attempts are developed. These scales include asking nicely, displaying anger, bargaining, showing affection, begging and pleading, just asking, and conning. These seven scales are shown to demonstrate adequate internal consistency, discriminant validity, and construct validity. Future research suggestions and scale applications are also discussed.

INTRODUCTION

Understanding children’s purchase influence has been identified as an area in great need of research (Harrigan 1991, Ward, Klees and Wackman 1990). In fact, Stipp (1993) states that "big gaps exist in our understanding of young people’s consumer behavior." The fact that knowledge gps exist is surprising given the importance of children in the marketplace. Changing social trends such as working mothers, decreased and delayed childbearing, increased divorce rates, and rapid maturation have given way to a formidable market forceBchildren. In 1997, children aged 4 to 12 were estimated to have a total income of approximately $24.4 billion, most of which was spent on products and services for immediate consumption (McNeal 1998). In addition to their personal income, children also influence allocations of expenditures in the household (Stipp 1993, Power et al. 1991). McNeal (1998) estimates that children influence spending in 62 product categories, equaling roughly $188 billion in purchases.

Given the aforementioned statistics, it would seem especially important to marketers to understand the process underlying how children influence family purchase decisions. However, with the exception of Isler, Popper and Ward’s (1987) model of children’s requests and parental responses, explanations of the child influence process are generally lacking in the literature. For this reason, a critical gap in the research on child influence is a conceptual explanation of how children influence purchase decisions. The present research begins to address this question by developing a scale based on social power theory to measure children’s direct influence attempts.

CHILD INFLUENCE RESEARCH

Research on the influence of children in family purchase decision-making dates back to the 1960s when Berey and Pollay (1968) conducted a study on the child’s role as influencer in cereal purchase decisions. Since that time, several studies have examined the degree to which parents yield to children’s influence attempts. Results of these studies revealed that parents were more likely to yield to children’s influence attempts when the purchase was for a product to be used by the child (Ward and Wackman 1972), when the child was female, or when the child was middle class (Atkin 1978). In addition, one study found that parents were more likely to yield to children’s influence attempts when the child was older (Atkin 1978).

In addition to parental yielding, a number of studies have assessed children’s influence on family decision processes. Across these studies, children exerted the most influence during problem recognition and search stages and the least influence in choice (Szybillo and Sosanie 1977, Nelson 1978, Belch et al. 1985). Further, children exerted little influence on the decisions of how much to spend, where to go, and transportation mode (Szybillo and Sosanie 1977, Jenkins 1979, Belch et al. 1985).

Numerous other studies have shown that child influence on family purchase decisions varies by product. For example, Foxman and Tansuhaj (1988) studied the impact of product category and product importance on the relative influence of family members in purchase decisions. They found that children have more influence for child products than for family products. Research has also indicated that children are influential in the purchases of cereal (Belch et al. 1985), vacations (Belch et al. 1985, Jenkins 1979), toys (Burns and Harrison 1985), and movies (Darley and Lim 1986).

A number of studies have examined how demographic variables specific to the child, such as age, gender, and income, affect the child’s influence in family decisions. It has been shown that children have more influence in purchase decision making as they grow older (Atkin 1978, Darley and Lim 1986, Moschis and Mitchell 1986, Nelson 1978, Ward and Wackman 1972). In addition to the child’s age, the gender of the child may affect his/her influence in family decisions. One study found that female children were more influential than male children, and children who earn income may have more influence in purchase decisions than those children who do not (Moschis and Mitchell 1986).

Other studies haveassessed the degree to which children’s influence varies by the child-rearing attitudes of the parents. Roberts et al. (1981) surveyed mothers and found that children had less influence when their mothers were more traditional or conservative. Darley and Lim (1986) examined the effect of locus of control of the parent on the child’s influence. They found that external locus of control parents perceived that the child had more influence across product categories and subdecisions than did internal locus of control parents. Berey and Pollay (1968) studied the child centeredness of parents and found that the more child-centered the mother was, the less likely she was influenced by the child.

Finally, studies have examined the degree to which children’s influence varies by demographic variables specific to the family unit. These studies examined the effects of social class, household income, family type, and family size. Results indicated that children who were members of middle class and higher income families had more influence in purchase decisions than children in low income, low social class families (Atkin 1978, Moschis and Mitchell 1986, Nelson 1978). In addition, children who lived in single parent households had greater relative influence than children in dual parent households (Darley and Lim 1986). Finally, in large families, children were more likely to be involved in decision-making processes (Nelson 1978).

In summary, studies have found that children exert varying degrees of influence on family decision processes and that children’s influence varies by product, child, parental and family characteristics. However, very little research has explored how children exert influence (for exception, see Isler et al. 1987). Theoretical development as to the process underlying children’s influence is limited. To begin to address this question, this research develops a scale based on social power theory of children’s direct influence attempts. It is the ultimate goal of this stream of research to include this measurement in a theoretically driven model of children’s relative influence.

SOCIAL POWER THEORY

Social power theory is a theoretical framework that examines power bases in social interactions. The concept of social power was initially introduced by Lewin (1951) and later developed theoretically by French and Raven (1959). Social power theory has been utilized as the conceptual framework in a number of marketing studies concerned with negotiation and conflict resolution. Two such studies in family decision-making used social power theory as a conceptual framework to underlie their research on influence attempts. Spiro (1983) sampled husbands and wives and developed a scale to measure influence strategies based on French and Raven’s five social power bases. Corfman and Lehmann (1987) utilized social power theory to develop a framework to explain relative influence in group decision-making.

Social power theory is also an appropriate theoretical framework from which to study children’s direct influence attempts for several reasons. First, social power theory identifies the bases of power which are present in social interactions. These sources of power may include expertise, reward power, referent power, legitimate power, and coercion (French and Raven 1959). These five bases are considered by French and Raven (1959) to be the most important resources utilized by a person, but not an all inclusive list of potential resources. Further, social power theory suggests that a person will make an assessment of his or her resources and choose an influence attempt that is consistent with his or her sources of social power (French and Raven 1959). In addition, social power theory suggests that these power bases may be utilized in two ways: active (i.e., direct) or passive (i.e., indirect). When the utilization of power to influence is active, or the result of an intentional act, a direct influence atempt is exerted.

CONSTRUCT DEFINITION AND DIMENSIONALITY

Following the conceptual framework of social power theory, children’s influence may be divided into two categories: direct (i.e. active) and indirect (i.e. passive) (Rossiter 1978). A direct influence attempt is defined as the agent’s actions which are intended to affect a change in behavior, attitude, goal, need or value on the part of the target. In parent-child interactions, a direct influence attempt may be considered the sum of a child’s actions intended to direct a decision outcome according to his or her own preferences. Unlike passive influence where a parent’s perception of a child’s unstated preferences influences a purchase decision (Wells 1965), a direct influence attempt encompasses only those instances where a child takes goal-directed action toward influencing a decision. In other words, a direct influence attempt is explicitly exerted and explicitly perceived. The scales developed in this research are intended to measure the direct or active influence attempts which children exert in a purchase situation.

Direct influence is multidimensional, with each dimension corresponding to different influence strategies. Few studies have examined the dimensionality of children’s direct influence attempts. Cowan and Avants (1988) and Cowan et al. (1984) extend Falbo and Peplau’s (1980) power strategies to children. Following an analysis of children’s written essays, Cowan et al. (1984) found the following fourteen strategy types: asking, begging and pleading, telling or assertion, reasoning, demanding or arguing, state importance, bargaining, persistence, negative affect, positive affect, verbal manipulation, eliciting reciprocity, using an advocate, evasion, and laissez-faire. Later, Cowan and Avants (1988) surveyed both mothers and children to identify influence strategies. Strategies identified and tested include ask, bargain, positive feelings, do as I please, tell, negative feelings, persistence, beg and plead, good deeds, reasoning, cry and get angry. In an early study of children’s interpersonal tactics, Wood et al. (1967) interviewed sixteen children with the objective of discovering children’s manipulation tactics. Results revealed five dimensions of strategies: norm invocation (appeals to rules, fair play, reason, etc.), positive sanctions (gifts, favors, bargaining, politeness, etc.), negative sanctions (physical aggression, nagging, begging, crying, etc.), ask, and don’t know or other.

In the purchase context, a gap exists in the empirical assessment of influence attempt dimensionality for children. However, a few studies do reference categorizations of influence strategies. Atkin (1978) alludes to asking and telling strategies when he summarizes children’s influence attempts in selecting a cereal purchase. McNeal (1992) refers to children’s request "styles" as those ways in which a child asks for something. He defines seven request styles: pleading, persistent, forceful, demonstrative, sugar-coated, threatening, and pity. Finally, Isler, Popper and Ward (1987) examined children’s purchase requests and parental responses via a diary study of 261 families. In their conceptual model, they refer to four request types: just ask, plead, bargain and other.

In summary, research has shown that direct influence attempt is multi-dimensional. Children employ a number of actions which are intended to cause a change in behavior, attitude, goal, need or value on the part of their target. Dimensions found across a number of the aforementioned studies include: asking, begging and pleading, bargaining, politeness, positive affect, manipulation, displaying anger, and crying and pouting.

SCALE DEVELOPMENT

Elicitation Procedure

Literature suggests that direct influence attempts are multi-dimensional, including several types of influence strategies. To further verify the multi-dimensional nature of children’s purchase influence, semi-structured in-depth interviews were conducted with 20 children aged 8 to 11 prior to item generation. In addition to the twenty children, twelve of the children’s mothers were interviewed over the telephone. Children were asked how they behaved when they wanted their parents to buy something for them. Mothers were also encouraged to relate past experiences when their children attempted to influence purchase behavior. The interviews revealed that children did employ various strategies to exert influence. The interviews, in addition to the strategies identified in aforementioned research, were used to identify a typology of children’s direct influence attempts. Table 1 offers a label, definition and example items for each strategy type.

Item Generation and Judgment

A pool of 128 items was generated from a review of the literature and from the interviews with 20 elementary school children and 12 of their mothers. These items were based on the definitions found in Table 1. Two faculty and three Ph.D. student judges evaluated the content validity of each item. Judges were provided with a definition of the overall construct of children’s direct influence attempt as well as the dimensions outlined in Table 1. Judges were asked to identify to which influence dimension each item best fit. In addition, each judge evaluated the degree to which each item was representative of the dimension to which it was assigned. Items which were classified in the correct dimension by at least four of the five judges and which received a representative rating of at least 3 on a 1=not at all representative to 5=very representative scale were retained for further analysis. This initial evaluation resulted in 106 items.

TABLE 1

DIRECT INFLUENCE ATTEMPT DIMENSIONS

Purification: Studies 1 and 2

In the first study, item analysis was performed on 106 items using a sample of 272 children aged 8-11. Items for each dimension were randomly listed throughout the questionnaire. Five point frequency scales derived from prior research (Ward and Wackman 1972) were used to measure the extent to which each child utilized each type of influence strategy. Response choices were "always", "a lot", "sometimes", "not much", and "never."

Iterative principal components factor analyses with oblique rotations were performed on the 106 items (Churchhill 1979, DeVellis 1991). Items which did not exhibit simple structure on factors or which loaded less than .50 were deleted. This process eliminated 39 items. The rotated factor pattern supported the designated dimensions "just ask," "bargain," "ask nicely," "show affection," "beg and plead," "show anger," "cry or pout," and "con." The eight factors were represented as follows: 13 items for ask nicely, 7 items for beg, 9 items for affection, 7 items for just ask, 12 items for anger, 4 items for con, 11 items for bargain, and 4 items for cry.

In the second study, the reliability and validity of the remaining 67 items were examined in a sample of 244 children aged 8-11. The responses were subject to iterative principal components factor analyses with oblique rotations. In the first analysis, the factor solution was constrained to eight factors. Examination of the eight factor solution found that all of the items reflecting the cry dimension cross-loaded very highly on other factors. Further examination of the frequency distribution of the cry items found that children reported a low frequency of use of this type of influence attempt. In addition, given the age range of the children in the sample, it is also probable that a social deirability bias may have influenced children’s responses about crying behavior. For these reasons, the factor analysis was unable to recover the cry factor. Thus, a seven factor solution was computed and refined, resulting in 47 items. These seven factors accounted for 56.2% of the variance. Coefficient alphas were computed for each scale. Items with item-to-total correlations of less than .5 were deleted from further analysis (Zaichowsky 1985). Five items were deleted. Forty-two items were retained.

In summary, principal components factor analyses and coefficient alphas were computed for the seven scales across two samples of children aged 8-11. Across these analyses the initial 106 items were reduced to 42 items across seven scales. These scales are "just ask," "bargain," "ask nicely," "show affection," "beg and plead," "show anger," and "con." The seven scales were represented by 7 items for ask nicely, 7 items for anger, 6 items for bargain, 9 items for affection, 7 items for beg, 3 items for ask, and 3 items for con.

Dimensionality and Internal Consistency: Studies 1 and 2

The 42 items retained were analyzed in confirmatory factor analyses via LISREL 8.12. A seven-factor model representing the hypothesized structure of the scale was estimated in order to assess discriminant validity and internal consistency.

In the first study, all items showed significant loadings on their factors. Overall fit of the seven-factor model, however, was attenuated due to items with extremely high correlations and error terms. For this reason, 13 problematic items were eliminated. This resulted in 6 items for the ask nicely scale, 5 items for the show affection scale, four items for the bargain, beg and plead, and show anger scales, and 3 items for the just ask and con scales. The final 29 items are included in the appendix.

The fit statistics and internal consistency estimates for both samples of the 29 item seven-factor models are presented in Tables 2 and 3. These fit statistics suggested adequate model fit for the seven-factor structure. The goodness-of-fit index (GFI) was .87 for both studies and the adjusted-goodness-of-fit index (AGFI) ranged from .84 to .85. In addition, Bentler’s (1990) comparative fit index (CFI) and the Tucker-Lewis index (TLI), which are robust to sampling characteristics, were .94 for both studies, which is within the acceptable range for designating adequate fit (Bentler 1990; Bollen 1989).

TABLE 2

FIT STATISTICS

TABLE 3

INTERNAL CONSISTENCY

Evidence for internal consistency is suggested by composite reliability, coefficient alpha, and variance extracted estimates (Fornell and Larcker 1981). Across the seven scales, composite reliability ranged from .63 to .91. Similarly, coefficient alpha ranged from .61 to .91. Variance extracted estimates assess the amount of variance which is captured by a construct’s measures relative to random measurement error. Variance extracted estimates ranged from .36 to .62. Finally, for all seven scales, all items had significant loadings on their constructs as reflected by significant t-values (p<.01), and the individual item reliabilities (i.e. the square of the standardized loading for each item) ranged from .26 to .74 across both studies.

Tests of discriminant validity were also performed on the seven factor model. First, the f estimates across samples ranged from .01 to .70. With one exception, [The f2 between bargain and affection in study 1 was equal to .49 and the average variance extracted between the two was .43.] all pairs of f2 were less than the average variance extracted between the two factors. Second, six-factor models were estimated and compared to the hypothesized seven-factor model. Each six-factor model combined the items of two scales into one overall factor and allowed the remaining factors to be separate but correlated. If the chi-square fit of the seven-factor model is better than the fit of the six-factor models, evidence of discriminant validity exists (Anderson and Gerbing 1988). Forty-two six-factor models across both studies were computed and compared to the seven-factor model. For all comparisons, the sevenfactor model was a better fit than the six-factor model (p<.001). These tests lend support for discriminant validity among the seven independent yet correlated scales (Anderson and Gerbing 1988; Fornell and Larcker 1981).

TABLE 4

CORRELATIONS AMONG DIRECT INFLUENCE ATTEMPT SCALES AND RELATED CONSTRUCTS: STUDY 1

Construct Validity: Studies 1 and 2

A number of measures were included in studies 1 and 2 for validity testing. For study 1, these measures included 10 items derived from the Narcissistic Personality Inventory (Raskin and Terry 1988) and 12 items derived from Social Power scales (Swasy 1979). All items were adjusted for use with children. The 10 items from the NPI were composed of 5 items for entitlement and 5 items for exploitativeness. The 12 items from Social Power were measured as parental power and were composed of 6 items for a parent’s reward power, 3 items for a parent’s legitimate power, and 3 items for a parent’s coercive power. In study 2, an additional 4 items which measure the child’s perception of his/herself as a polite person were included. All measures were scored on four point Likert-type scales designed for use with children.

Results of the correlations between the entitlement, exploitativeness, reward power, legitimate power, coercive power and politeness scales and the seven direct influence attempt scales are depicted in Tables 4 and 5. The correlations support the construct validity of the seven direct influence attempt scales.

Entitlement is designed to measure aspects of narcissism where a person expects special favors or privileges over others without obligation for reciprocity (Raskin and Terry 1988). Children who feel that they deserve to get their way are likely to resort to deception, to display anger, or to beg and plead when attempting to exert influence. [Note that the types of direct influence attempts which are highly correlated with entitlement are in strong contrast to the types of direct influence attempts which are highly correlated with legitimate power. This further supports the notion that when children feel that they deserve to have their way, they react by utilizing negative influence attempts (e.g., displaying anger, con, and begging and pleading). In contrast, when children feel that their parent has the right to tell them what to do, they react by utilizing positive influence attempts (e.g., asking nicely, showing affection, and bargaining).] Thus, entitlement should positively correlate with the con, anger, and beg and plead scales. As Tables 4 and 5 suggest, a positive correlation is found between entitlement and con and anger in both studies (p<.01). In study 1, entitlement is positively correlated with beg and plead (p<.01).

Exploitativeness is designed to measure the aspects of narcissism where a person feels that he/she can understand others and thereby manipulate them. Con reflects this sort of manipulation, where a child feels he/she is able to get his/her way by deceiving the parent. Additional direct influence attempt scales which may be associated with exploitativeness are anger and beg and plead. As depicted in Tables 4 and 5, a positive correlation exists between exploitativeness and con in both studies (p<.01). In study 1, a positive correlation exists between exploitativeness and beg and plead (p<.01). In study 2, a positive correlation exists between exploitativeness and anger (p<.01).

Coercive power resides in the other parties’ perception that punishment will result from non-compliance (French and Raven 1959). As measured in this study, coercive power is the degree to which children believe that they must do as their parents request in order to avoid punishment. Children who fear punishment should attempt to exert influence in ways which will be perceived as positive by parents. For this reason, coercive power should be positively correlated to asking nicely, showing affection, and bargaining. Tables 4 and 5 support this suggestion, as coercive power is positively correlated with ask nicely, affection and bargain in both samples (p<.05).

Legitimate power is the degree to which a person is perceived to have the right to exert influence (French and Raven 1959). It represents the degree to which children perceive that their parents have the right to tell them what to do. Children who perceive that their parents have legitimate power should not attempt to exert influence by displaying anger, begging, or attempting to con their parents. In contrast, children should behave in a favorable manner toward their parents, suggesting that children will attempt to exert influence by asking nicely, showing affection, and bargaining. As suggested by Tables 4 and 5, this pattern of results is supported. Ask nice is positively corrlated with legitimate power in both studies (p<.01). In study 1, affection and bargaining are positively correlated with legitimate power (p<.05) and anger is negatively correlated with legitimate power (p<.05). In study 2, beg and con are negatively correlated with legitimate power (p<.05 and p<.01).

TABLE 5

CORRELATIONS AMONG DIRECT INFLUENCE ATTEMPT SCALES AND RELATED CONSTRUCTS: STUDY 2

Reward power is the ability to positively or negatively bestow something on another party (French and Raven 1959). In this context, reward power is the degree to which children comply with parents’ requests in order to extract a reward. In other words, children express compliance in exchange for some future gain. Reward power should be positively correlated with the bargain scale and the con scale. As shown in Tables 4 and 5, reward is positively correlated with bargain (p<.01) and con (p<.05) in both studies.

The child’s perception of his/herself as a nice or polite person was measured with four items, "I am a polite person," "I generally try to be nice to other people," "I am a nice person, " and "I try to be a sweet person." A child’s assessment of the degree to which they perceive themselves to be nice or polite should be positively correlated with the asking nicely and showing affection scales and negatively correlated with the anger scale. As shown in Table 5, nice is positively correlated with ask nicely and affection (p<.01) and negatively correlated with anger (p<.05).

SUMMARY AND FUTURE RESEARCH

Scales measuring seven dimensions of children’s direct influence attempts were developed and validated. The scale development procedure included both qualitative and quantitative data collection phases. Three studies with a total of 536 children demonstrated the structure of children’s direct influence attempts. The seven dimensions derived to represent direct influence attempt were ask nicely, bargain, show affection, just ask, beg and plead, show anger, and con.

The development of a direct influence attempt scale suggests several additional avenues for future research. For example, at present, it is unknown what proportion of influence attributed to children is derived from overt action or parental concession. As noted by McNeal (1992), the distinction between active and passive influence is an area in need of explication. This research provides the first step in such a distinction by developing measures for direct influence attempt. With the development of a direct influence attempt scale, the conditions under which children employ the various influence attempts can be investigated.

Further, to complement the direct influence attempt measure, the next step in this stream of research is the development of an indirect influence attempt scale. In addition to developing measures, children’s direct and indirect influence should be tested in a more comprehensive conceptual model of children’s relative influence. In addition to social power theory, several theoretical frameworks may be used as frameworks from which to support such a model. Theories which could be investigated in future research include social exchange theory, social comparison theory, role theory, and resource theory.

Finally, future research should also explore the determinants of children’s direct influence attempts. Children’s direct influence attempt scales were shown to be significantly related to (yet distinct from) NPI’s entitlement and exploitativeness and the social power scales reward, coercion, and legitimate. This lends support to the notion that children are able to assess their personal resources and choose an appropriate influence attempt for the situation. Therefore, it seems likely that personal resources are a primary determinant of children’s direct influence attempts.

APPENDIX

REFERENCES

Anderson, James C. and David W. Gerbing. (1988) "Structural Equation Modeling in Practice: A Review and Recommended Two-Step Approach," Psychological Bulletin, 103 (November), 411-423.

Atkin, Charles K. (1978) "Observation of Parent-Child Interaction in Supermarket Decision-Making," Journal of Marketing, 42 (October), 41-45.

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

Bentler, Paul M. (1990) "Comparative Fit Indices in Structural Equation Models," Psychological Bulletin, 107 (March), 238-246.

Berey, Lewis A. and Richard W. Pollay. (1968) "The Influencing Role of the Child in Family Decision Making," Journal of Marketing Research, 5 (February), 70-72.

Bollen, Kenneth A. (1989) Structural Equations with Latent Variables. New York: Wiley.

Burns, Alvin C. and Mary C. Harrison. (1985) "Children's Self-Perceptions of Their Participation in Retail Store Patronage Decisions," Advances in Consumer Research, 12.

Churchhill, Gilbert A., Jr. (1979) "A Paradigm For Developing Better Measures of Marketing Constructs," Journal of Marketing Research, 16 (February), 64-73.

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, 14 (June), 1-13.

Cowan, Gloria and S. Kelly Avants. (1988) "Children's Influence Strategies: Structure, Sex Differences, and Bilateral Mother-Child Influence," Child Development, 59, 1303-1313.

Cowan, Gloria, Joan Drinkard, and Laurie MacGavin. (1984) "The Effects of Target, Age, and Gender On Use of Power Strategies," Journal of Personality and Social Psychology, 47 (6), 1391-1398.

Darley, William F. and Jeen-Su Lim. (1986) "Family Decision Making in Leisure-Time Activities: An Exploratory Analysis of the Impact of Locus of Control, Child Age Influence Factor and Parental Type of Perceived Child Influence," Advances in Consumer Research, 13, 370-374.

DeVellis, Robert F. (1991) Scale Development: Theory and Applications. Newbury Park, CA: Sage.

Falbo, Toni and Letitia Anne Peplau. (1980) "Power Strategies in Intimate Relationships," Journal of Personality and Social Psychology, 38(4), 618-628.

Fornell, Claes and David F. Larcker. (1981) "Evaluating Structural Equation Models with Unobservable Variables and Measurement Error," Journal of Marketing Research, 18 (February), 39-50.

Foxman, Ellen R. and Patriya S. Tansuhaj. (1988) "Adolescents' and Mothers' Perceptions of Relative Influence in Family Purchase Decisions: Patterns of Agreement and Disagreement," Advances in Consumer Research, 15, 449-453.

French, John R. P. and Bertram Raven. (1959) "The Basis of Social Power," Studies in Social Power, Cartwright, ed. Ann Arbor, MI: Institute for Social Research.

Harrigan, Judy A. (1991) "Children's Research: Where It's Been, Where It Is Going," Advances in Consumer Research, 18, 11-17.

Isler, Leslie, Edward T. Popper and Scott Ward. (1987) "Children's Purchase Requests and Parental Responses: Results from a Diary Study," Journal of Advertising Research, October / November, 28-39.

Jenkins, Roger L. (1979) "The Influence of Children in Family Decision-Making: Parents' Perceptions," Advances in Consumer Research, 6, 413-418.

Lewin, Kurt. (1951) Field Theory in Social Science. New York: Harper.

McNeal, James U. (1992) "The Littlest Shoppers," American Demographics, February, 48-53.

McNeal, James U. (1998) "Taping the Three Kids' Markets," American Demographics, 20 (4), 36-41.

Moschis, George P. and Linda G. Mitchell. (1986) "Television Advertising and Interpersonal Influences on Teenagers' Participation in Family Consumer Decisions," Advances in Consumer Research, 13, 181-186.

Nelson, James E. (1978) "Children as Information Sources in Family Decision to Eat Out," Advances in Consumer Research, 6, 419-423.

Power, Christopher, Sandra Atchison, Gail DeGeorge and Dean Foust. (1991) "Getting 'Em While They're Young," Business Week, September 9, 94-95.

Raskin, Robert and Howard Terry. (1988) "A Principal Components Analysis of the Narcissistic Personality Inventory and Further Evidence of Its Construct Validity," Journal of Personality and Social Psychology, 54 (May), 890-902.

Roberts, Mary Lou, Lawrence H. Wortzel and Robert L. Berkeley. (1981) "Mothers' Attitudes and Perceptions of Children's Influence and Their Effect on Family Consumption," Advances in Consumer Research, 8, 730-735.

Spiro, Rosann L. (1983) "Persuasion in Family Decision-Making," Journal of Consumer Research, 9 (March), 393-402.

Stipp, Horst. (1993) "New Ways to Reach Children," American Demographics, August, 50-56.

Swasy, John L. (1979) "Measuring the Bases of Social Power," Advances in Consumer Research, 6, 340-346.

Szybillo, George J. and Arlene Sosanie. (1977) "Family Decision Making: Husbands, Wives and Children," Advances in Consumer Research, 4, 46-49.

Ward, Scott, Donna M. Klees and Daniel B. Wackman. (1990) "Consumer Socialization Research: Content Analysis of Post-1980 Studies and Some Implications for Future Work," Advances in Consumer Research, 17, 798-803.

Ward, Scott and Daniel B. Wackman. (1972) "Children's Purchase Influence Attempts and Parental Yielding," Journal of Marketing Research, 9, 316-319.

Wells, William D. (1965) "Communicating With Children," Journal of Advertising Research, 5, 2-14.

Wood, James R., Eugene A. Weinstein, and Ronald Parker. (1967) "Children's Interpersonal Tactics," Sociological Inquiry, 37 (Winter), 129-138.

Zaichkowsky, Judith Lynne. (1985) "Measuring the Involvement Construct," Journal of Consumer Research, 12 (December), 341-352.

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