Gender Differences in Information Processing Confidence in an Advertising Context: a Preliminary Study

DeAnna S. Kempf, Iowa State University
Kay M. Palan, Iowa State University
Russell N. Laczniak, Iowa State University
ABSTRACT - Previous research suggests that males and females process advertising information differently. However, researchers have yet to explore whether biological sex influences the confidence that receivers have in their ad responses. In addition, prior research has failed to determine the relative influence of biological sex and gender identity on confidence. Results from an experiment featuring a fictitious soft drink brand suggest that males exhibit greater confidence in their attitudes toward the non-claim component of an ad, and suggest that gender identity accounts for variance in several different measures of advertising processing confidence, beyond what was explained by biological sex.
[ to cite ]:
DeAnna S. Kempf, Kay M. Palan, and Russell N. Laczniak (1997) ,"Gender Differences in Information Processing Confidence in an Advertising Context: a Preliminary Study", in NA - Advances in Consumer Research Volume 24, eds. Merrie Brucks and Deborah J. MacInnis, Provo, UT : Association for Consumer Research, Pages: 443-449.

Advances in Consumer Research Volume 24, 1997      Pages 443-449

GENDER DIFFERENCES IN INFORMATION PROCESSING CONFIDENCE IN AN ADVERTISING CONTEXT: A PRELIMINARY STUDY

DeAnna S. Kempf, Iowa State University

Kay M. Palan, Iowa State University

Russell N. Laczniak, Iowa State University

ABSTRACT -

Previous research suggests that males and females process advertising information differently. However, researchers have yet to explore whether biological sex influences the confidence that receivers have in their ad responses. In addition, prior research has failed to determine the relative influence of biological sex and gender identity on confidence. Results from an experiment featuring a fictitious soft drink brand suggest that males exhibit greater confidence in their attitudes toward the non-claim component of an ad, and suggest that gender identity accounts for variance in several different measures of advertising processing confidence, beyond what was explained by biological sex.

INTRODUCTION

Advertising researchers are becoming increasingly concerned with the impact of gender differences on consumer responses to advertising. Indeed, several studies have reported significant differences between males and females in the procssing of advertising information (e.g., Darley and Smith 1995; Meyers-Levy and Maheswaran 1991). What researchers have not yet examined, however, are gender differences with respect to information processing confidence. Specifically, do males and females have different levels of confidence when processing information contained in advertising? Several studies have reported that males and females differ with respect to self-confidence (Lenney 1977; Maccoby and Jacklin 1974; White, DeSanctis, and Crino 1981), suggesting that differences in information processing confidence might also be present in an advertising context. One purpose of this paper is to conduct a preliminary examination of the effect of biological sex, i.e., whether an individual is male or female, on generalized information processing confidence and belief and attitude confidence resulting from an advertising exposure. Further, since there is evidence that gender identity, i.e., an individual’s gender self-concept, may explain differences in consumers’ responses to advertising (e.g., Coughlin and O’Connor 1985; Jaffe 1991), a secondary purpose is to explore the ability of gender identity to explain differences in information processing confidence and belief and attitude confidence beyond that explained by biological sex alone.

CONFIDENCE CONSTRUCTS IN ADVERTISING RESPONSE

Evaluative confidence, defined as the certainty with which an evaluative judgment is held, is applicable to any consumer evaluation or judgment, but has been studied by marketing and consumer researchers mainly in the form of belief confidence (ci in Fishbein’s expectancy value model of attitudes) and brand attitude confidence (e.g., Bennett and Harrell 1975; Fishbein and Ajzen 1975; Smith and Swinyard 1982; 1983; 1988). Bennett and Harrell (1975) and Smith and Swinyard (1988) found that inclusion of the belief confidence measure as an indication of how certain the consumer feels about his/her belief probability estimate improves the predictive value of the Fishbein and Ajzen model. Researchers have long desired to accurately predict behavior from measured attitudes, and attitude confidence is one variable that has been shown to have significant positive effects on attitude and behavior consistency (Fazio and Zanna 1978; Smith and Swinyard 1983). Therefore, a deeper understanding of the antecedents of confidence would be valuable to both theory and practice in the attitude area. This study makes a first step toward understanding two related antecedents of confidence in a marketing contextCbiological sex and gender identity.

Although evaluative confidence has been studied in marketing primarily with respect to brand beliefs and brand attitudes, it is also relevant for evaluations of the ad itself (AAd), including both claim and non-claim components (Miniard, Bhatla, and Rose 1990). In this study, therefore, the effects of biological sex and gender identity on belief confidence, attitude confidence, and attitude-toward-the-ad confidence are investigated.

Marketing researchers have hypothesized numerous effects of confidence, with mixed results. According to Smith and Swinyard (1988), this lack of consistency in empirical findings involving belief and attitude confidence is because most studies including confidence have not specifically manipulated antecedent conditions that should create significant variation in subjects’ confidence levels. One exception is the research stream dealing with the effects of product trial, which has shown that attitudes and beliefs formed as a result of a direct trial experience are more confidently held than those formed from advertising or other indirect sources of information (e.g., Fazio and Zanna 1978; Smith and Swinyard 1988; Smith 1993).

Several antecedents of confidence have been proposed including: (1) the quantity of information available for the judgment (e.g., Dover and Olson 1977; Peterson and Pitz1988), (2) the credibility of the information (Fishbein and Ajzen 1975), and (3) the consistency of the information among sources (Kahneman and Tversky 1973). The current study examines the potential effects of two additional and related variables on confidence constructs: biological sex and gender identity.

While the confidence constructs of belief confidence and attitude confidence discussed above have been studied frequently in the advertising literature, another constructCgeneralized information processing confidence (GIPC)Chas received little attention (Wright 1975). GIPC refers to the degree of certainty or confidence an individual feels in his/her ability to form judgments in response to information exposure. Wright (1975) defined high GIPC more generally as "having quick mental reflexes," although it is important to note that in that study, as well as the current one, GIPC refers to the consumer’s self-perceived level of ability rather than an objective measure of ability to process information. Again, as with the various forms of evaluative confidence discussed above, the current study investigates the effects of biological sex and gender identity on GIPC.

Although sex and gender identity have not been studied with respect to generalized information processing confidence measures, previous research indicates that both variables are significantly related to another type of person-level confidenceCself-confidence. Maccoby and Jacklin (1974) reported that self-confidence is lower for females than for men in almost all achievement situations. In a separate review of the literature, Lenney (1977) concluded that self-confidence among women was lower than among men in only three types of achievement situations: when the specific task being measured is sex-linked, when clear and unambiguous performance feedback is unavailable, and when it is suggested that females’ work will be compared to or evaluated by others. More recent studies support Lenney’s (1977) contention that differences in self-confidence are situational. For example, Zuckerman (1985) found no differences among female and male students’ perceptions of interpersonal self-confidence, but Andrews (1987) found that females had lower self-confidence than males concerning their ability to communicate arguments persuasively. However, Chusmir, Koberg, and Stecher (1992) found no differences in self-confidence among males and females in either the work setting or in social/family situations. In the same study, though, gender identity was significant with respect to self-confidence. Both males and females scoring high on masculine attributes reported greater self-confidence in work than in social situations; the same result also occurred for both males and females scoring high on feminine attributes. Based on these studies, it is reasonable to expect that both biological sex and gender identity will be related to information processing confidence.

GENDER IDENTITY AND BIOLOGICAL SEX IN ADVERTISING RESPONSE

Gender identity represents an individual’s self-perceived endorsement of masculine and feminine personality traits, and as such, may or may not be congruent with an individual’s biological sex. Therefore, gender identity potentially has effects distinct from those related to biological sex. However, whereas biological sex generally has played a significant explanatory role in consumer research, gender identity has rarely been found to be a significant predictor (Stern 1988), except when used to explain behaviors that are clearly gender-related (Fischer and Arnold 1994). Yet, it is reasonable to expect gender identity to be significant in explaining differences in information processing confidence since Chusmir et al. (1992) found significant relationships between gender identity characteristics and self-confidence. Moreover, biological sex was not significantly related to self-confidence in the same study, suggestingthe effects of biological sex and gender identity may be different with respect to information processing confidence.

While several different conceptualizations of gender identity exist, this paper takes the position that gender identity is a multidimensional construct where masculinity and femininity are orthogonal dimensions, coexisting in varying degrees within an individual (Bem 1974, 1975; Spence, Helmreich, and Stapp 1975). Typically, an individual is placed into one of four different categories based on his/her endorsement of both masculine and feminine characteristics: (1) sex-typed, when an individual primarily endorses same-sex traits; (2) cross sex-typed, when an individual primarily endorses opposite-sex traits; (3) androgynous, when an individual endorses both masculine and feminine characteristics to a high degree; and (4) undifferentiated, when an individual endorses both feminine and masculine characteristics to a low degree (Spence et al. 1975). However, some researchers, concerned with the validity of combining sex-typed masculine males and feminine females together, have chosen instead to examine individuals who endorse masculine characteristics to a high degree, relative to feminine characteristics, as masculine schematics and individuals who endorse feminine characteristics to a high degree, relative to masculine characteristics, as feminine schematics (Gentry and Haley 1984; Markus, Crane, Bernstein, and Siladi 1982).

While somewhat limited, advertising research that has used either of these classification schemes for gender identity has produced interesting results. For example, Gentry and Haley (1984), using three different operationalizations of information processing, found only one significant relationship (out of fifteen possible relationships) between gender identity and ad recall. Coughlin and O’Connor (1985), examining the hypothesis that an individual’s reaction to sex-related role portrayals in ads would be consistent with his/her gender identity, found several significant relationships. As expected, androgynous males and females reacted similarly to ads, but, unexpectedly, cross sex-typed males and females responded unfavorably to non-traditional roles. Along these same lines, Jaffe and Berger (1988) and Jaffe (1991) found that cross sex-typed females (i.e., masculine females) and androgynous females prefer modern positioning of females in advertisements, while sex-typed females (i.e., feminine females) and undifferentiated females responded similarly to both traditional and modern positionings.

Biological sex has been shown to be significantly related to several aspects of consumer behavior important to advertising. According to the selectivity model of information processing (Meyers-Levy 1989), males tend to use a selective, heuristic mode of processing information suggesting they tend to base judgments on a single cue. Females, on the other hand, are more likely to use a comprehensive processing mode, basing judgments on all available cues. However, in situations where consumers are motivated to engage in detailed processing, males will drop schema-based processing in favor of detailed processing (Meyers-Levy and Maheswaran 1991). Darley and Smith (1995), using the selectivity model, found significant differences in male and female responses to objective and subjective ad claims. Biological sex has also been found to be significant in response to "gendered ads," i.e., ads specifically designed to appeal to either men or women. For instance, Alreck, Settle, and Belch (1982) found that men were more likely to try and use a masculine brand of soap and women were more likely to try and use a feminine brand of soap, but that masculine gendering was generally more acceptable than feminine gendering to all respondents. Gentry and Doering (1977) found biological sex to be significant in predicting attitudes toward and usage of gendered products.

Importantly, studies that have examined biological sex and gender identity together have generally found that biological sex is a better predictor than is gender identity (Alreck et al. 1982; Gentry and Doering 1977). In our study, both biological sex and geder identity will be examined. Gender identity will be represented in two ways: first, a categorization system for gender identity is used that combines elements from the traditional classification system (Bem 1974; Spence et al. 1975) with Markus et al.’s (1982) classification system, since consumer research has reported better results using the Markus et al. System (Gentry and Haley 1984). Consequently, respondents are organized into one of four gender identity categoriesCmasculine schematic, feminine schematic, androgynous, or undifferentiated. Second, since Spence et al. (1975) showed that identification with masculine characteristics was most strongly related to self confidence (compared to identification or non-identification with feminine characteristics), masculine characteristics are included in the analyses separately by summing the original 5-point scales over the eight masculine items in the PAQ scale, thereby approximating a continuous measure of identification with masculine characteristics.

HYPOTHESES

The previous discussion suggests several hypotheses. First, it is expected that both biological sex and gender identity will have direct effects on generalized information processing confidence (GIPC). Specifically, since male and female differences exist in information processing and in self-confidence, differences should also be present in GIPC. Further, since high self-confidence is associated with a masculine self-concept (Spence et al. 1975) and masculine traits (Chusmir et al. 1992), it is expected that the GIPC of individuals endorsing a high degree of masculine traits would be higher than for those individuals who do not endorse a high degree of masculine traits.

H1: Males will score higher on the GIPC relative to females.

H2: Masculine schematics and androgynous individuals will score higher on the GIPC relative to feminine schematics and individuals who are undifferentiated.

Second, it is expected that both biological sex and gender identity will directly affect several confidence constructs related to subjects’ responses to an ad exposure, specifically belief confidence, attitude confidence, AAd-claim confidence, and AAd-nonclaim confidence. While the hypothesized effects of biological sex are identical across these four dependent variables, since this is a first attempt at studying these effects, they are studied separately in order to highlight any differences in the observed pattern of effects, leading to the following hypotheses.

H3a: Males will report higher belief confidence relative to females.

H3b: Males will report higher attitude confidence relative to females.

H3c: Males will report higher AAd-claim confidence relative to females.

H3d: Males will report higher AAd-nonclaim confidence relative to females.

Spence et al. (1975) found that it was specifically subjects’ level of identification with masculine characteristics that had the highest correlation with self esteem; Chusmir et al. (1992) found a similar relationship between masculine characteristics and self-confidence. By extension, then, stronger identification with masculine characteristics is predicted to produce higher levels of confidence relating to the consumer’s advertising responses. Again, the four advertising-response confidence variabls are included in separate hypotheses.

H4a: Masculine characteristics will be positively related to higher belief confidence.

H4b: Masculine characteristics will be positively related to higher attitude confidence.

H4c: Masculine characteristics will be positively related to higher AAd-claim confidence.

H4d: Masculine characteristics will be positively related to higher AAd-nonclaim confidence.

METHOD

Procedure

Subjects for this study were undergraduate and graduate students enrolled in marketing courses at a medium-sized midwestern university. The final sample size was 105 students. The use of a student sample was deemed appropriate, since the use of a homogeneous sampling frame is recommended when a theory application test is being conducted (Calder, Phillips and Tybout 1982). Moreover, since the ad and product used in this study were targeted at college students, a student sample is appropriate.

In classroom settings, subjects were given a portfolio of three professionally developed ads (one test ad and two "filler" ads). The test ad featured a fictitious soft-drink brand (Citrus Springs), and included a picture of a model and superiority claims regarding the taste of the drink (determined to be the most salient attribute in a pretest of subjects similar to those used in the main study). A soft drink was chosen as the advertised product because students constitute a sizeable market for the product, soft drinks are consumed by both males and females, and use of such a product was consistent with the experimental guise utilized in the present study (i.e., subjects would be asked for their opinions of three new brands of products that might be of interest). Fictitious brands of virus protection software and 35mm cameras were featured in the filler ads. The use of unfamiliar brand names was deemed appropriate to ensure that prior brand knowledge and preferences would not influence post-exposure beliefs and attitudes.

Subjects were given two minutes to look through the ad portfolio (determined in a pretest to be sufficient to allow them to thoroughly view all ads). The mean score for a six-item measure of self-perceived advertising message involvement (Andrews and Shimp 1990Calpha=.89) was 4.43 on a seven point likert-type scale (where 1 indicated that subjects would strongly disagree that they were not involved with the test ad’s message and 7 indicates they would strongly agree that they were involved). This result indicates that subjects were moderately involved with the test ad’s message. Further, results suggested subjects were relatively homogeneous in their self-perceived ad message involvement. Print ads were used to allow subjects an adequate opportunity to elaborate on the test ad’s message so that they would be able to formulate beliefs and attitudes which could be confidently held.

After ad portfolios were collected, subjects were presented with questionnaires which gathered their cognitive responses to the test ad, self reports of ad message involvement, and measures of brand beliefs and brand-belief confidence, attitude toward the ad and attitude-toward-the-ad confidence, attitude toward the brand and attitude-toward-the-brand confidence, generalized information-processing confidence and gender identity. As a-priori hypotheses were not developed regarding several of these measures (i.e., cognitive responses, brand beliefs, attitude toward the ad, and attitude toward the brand), our discussion of study measures will focus only on a subset of these measures.

Measures

Brand-Belief Confidence. Belief confidence was measured by asking subjects how certain and confident they were regarding their reported taste beliefs for the advertised brand (Citrus Springs). Seven-point bipolar adjective scales with endpoints, "Not at all certain" and "Very certain" and "Not at all confident" and "Very confident," were used. An average of the six items was computed (for the three belief statements regarding taste) and used as the belief confidence measure (alpha=.85)

Attitude-toward-the-ad Confidence. As the present study views attitude toward the ad in a manner consistent with Miniard, Bhatla and Rose’s (1990) contention that attitude toward the ad needs to be decomposed into claims and non-claims components, two measures of attitude-toward-the-ad confidence were gathered. Subjects were asked their degree of certainty and confidence in their attitudes toward the ad’s claims (alpha=.98) and attitudes toward the ad’s overall appearance (alpha=.95) on 7-point bipolar adjective scales (with endpoints, "Not at all certain" and "Very certain" and "Not at all confident" and "Very confident").

TABLE 1

MEAN LEVELS OF GIPC BY GENDER IDENTITY

Attitude-toward-the-brand Confidence. Attitude-toward-the-brand confidence was assessed by asking subjects to rate their degree of certainty and confidence toward their brand attitudes with 7-point bipolar adjective scales with endpoints, "Not at all certain" and "Very certain" and "Not at all confident" and "Very confident." Alpha for this measure was .95.

It is important to note that all the above confidence measures were adapted from a previously published study which investigated effects on advertising on receivers’ confidence (Laczniak and Muehling 1993).

Generalized Information Processing Confidence. The generalized-information-processing confidence measure was derived from Wright’s (1975) study dealing with individual differences and information processing. This concept was measured with seven nine-point bipolar adjective items (where a response of 1 indicated that subjects felt a statement was "definitely true" and a score of 9 indicated the belief that a statement was "completely false"). Cronbach alpha for this measure was .79. Sample items include: "I often have trouble concentrating" and "I am certainly able to think quickly."

Gender Identity. Gender identity was measured using the Personal Attributes Questionnaire (PAQ) developed by Spence et al. (1975). While the PAQ has been infrequently used in consumer research, there is evidence that it may perform better than the Bem Sex Role Inventory when the moderating role of gender identity is examined (Palan, Areni, and Kiecker 1994). The PAQ contains 24 scale itemsC8 masculinity items, 8 femininity items, and 8 items that are non-sex specific. Coefficient alphas for the masculine and feminine scales were .83 and .80, respectively. For some of the analyses, these characteristics were summed and treated as continuous measures. For other analyses, median splits of the summed responses for the masculine and feminine scales were used to classify respondents as masculine schematic (high masculine, low feminine), feminine schematic (high feminine, low masculine), androgynous (high masculine and high feminine), or undifferentiated (low masculine and low feminine). This classification system resulted in 11 females and 15 males being categorized as masculine schematic, 10 females and 10 males categorized as feminine schematic, 14 females and 14 males categorized as androgynous, and 16 females and 15 males categorized as undifferentiated.

RESULTS

Effects of Sex and Gender Identity on Generalized Information Processing Confidence

One-way ANOVA results showed that males reported significantly higher levels of GIPC than females (F1,112=6.87, p<.01), supporting H1. The mean GIPC score (on a 9-point scale) for males was 6.54, and 5.89 for females. This result, as suspected based on previous research, clearly shows that males tend to possess higher information processing confidence than females.

Next, the effects of gender identity on GIPC were assessed via ANOVA. The results indicated that gender identity had a significant main effect on GIPC ( F3,99=18.70, p<.0001). The cell means for gender identity are shown in Table 1. Consistent with H2, masculine schematics and androgynous individuals had significantly higher GIPC scores than did feminine schematics and undifferentiated individuals.

To further investigate the effects of gender identity and sex on GIPC, hierarchical regression analysis (Cohen and Cohen 1983) was performed with sex as a dummy variable (F=1, M=0), and the summed scores for masculine and feminine characteristics included separately in the model as continuous variables. For the baseline model including only the sex dummy variable, the R2=.049. When masculine and feminine traits were added to the model, the R2 jumped dramatically to .47, a highly significant difference (Partial F test: F2,110=47.58, p<.0001). The two gender trait independent variables were statistically significant and positive (masculine traits: ¯=.62, t=8.71, p-.0001; feminine traits: ¯=.15, t=2.00, p-.041). The coefficient of the sex dummy variable was also statistically significant (¯=-.17, t=2.29, p-.02), indicating that males reported higher GIPC than females. The results of these analyses indicate that gender identity explains variance in GIPC above and beyond that explained by biological sex. And, as was found by Chusmir et al. (1992) with respect to self-confidence, identification with both masculine and feminine traits seems to increase GIPC, with androgynous individuals reporting the highest levels. Thus, H2 is supported, even with the effects of biological sex partialed out.

The second set of hypotheses predicts effects of gender identity and sex on confidence variables more commonly reported in advertising response research: (1) belief confidence, (2) attitude confidence, (3) AAd-claim confidence, and (4) AAd-nonclaim confidence. For the two-level sex variable, one-tailed t-tests were performed on the mean differences in these dependent variables. The results are shown in Table 2.

TABLE 2

MEANS OF CONFIDENCE VARIABLES ACROSS LEVELS OF SEX

TABLE 3

REGRESSION RESULTS FOR GENDER IDENTITY

As can be seen in Table 2, although the mean differences are in the hypothesized direction, only the mean differences for AAd-nonclaim confidence were statistically significant. Thus, H3d was supported, H3a-c were not. Apparently, males have significantly greater confidence than females only with respect to AAd-nonclaim confidence.

H4a-d were tested via regression analysis, with gender identity included in the model as the summed scores of masculine and feminine characteristics. The standardized regression coefficients and R2 values are shown in Table 3 below. As the results in Table 3 show, identification with more masculine traits led to higher levels of confidence except for belief confidence.

It is clear from the hypothesis tests that gender identity and biological sex do directly influence GIPC and the other confidence measures. To explore the possibility of a mediating relationship, a regression analysis was performed with the belief confidence, AB confidence, and AAd confidence measures regressed on GIPC. None of these relationships were significant, thus precluding a mediating relationship.

DISCUSSION

Importantly, this study presents initial evidence that both biological sex and gender identity are significant antecedents of GIPC. In addition, while biological sex was a sigificant antecedent of AAd-nonclaim confidence, gender identity proved to be an antecedent of attitude confidence, AAd-claim confidence, and AAd-nonclaim confidence. These results are consistent with previous research that has reported significant differences between males and females in information processing (e.g., Meyers-Levy and Maheswaran 1991), self-confidence (e.g., Maccoby and Jacklin 1974), and self-confidence related to gender identity (Chusmir et al. 1992). This illustrates the relative explanatory power of gender identity as compared to biological sex when studying confidence in an advertising context, a finding contrary to most consumer research that reports biological sex as having more explanatory power than gender identity. These findings suggest that whenever confidence measures are being studied, the effects of both gender identity and biological sex should be considered. At a minimum, these effects should be dealt with as covariates when they are not a central focus of the study.

For advertising practice, these results may indicate that for ads directed toward females, a single ad exposure may not be as successful at creating confidently-held attitudes as it would be for males. To explore this possibility further, future research should examine how additional ad exposures affect males’ and females’ confidence levels.

Limitations and Suggestions for Future Research

As with any research, several limitations, many of which could be addressed through future research, may have affected the results. First, the limited sample size may have hampered our ability to find significant results with respect to gender identity; larger sample sizes should be obtained in future studies. Second, our sample was relatively homogeneous in their self-assessments of ad message involvement. Future research may wish to manipulate this concept to determine if sex and gender identity differences noted occur for both more and less involved subjects. Third, because this was a student sample, there should always be caution in generalizing the results to the general population. Fourth, the ad upon which the confidence measures were based included very little copy; future studies should replicate this study using ads with differing levels of central route information to determine the robustness of the results obtained here. Examining the confidence variable more in-depth would also be valuable; for example, it may be that feminine schematics possess greater confidence in their beliefs and attitudes when recalling ads with strong emotional overtones, whereas masculine schematics have greater confidence regarding claims that can be counterargued. Finally, it might also be interesting to replicate the study using gendered ads to determine if an interaction effect would occur between the gendered ads, biological sex, and gender identity.

REFERENCES

Alreck, Pamela L., Robert B. Settle, and Michael A. Belch (1982), "Who Responds to "Gendered" Ads, and How?" Journal of Advertising Research, (April/May), 25-32.

Andrews, J. Craig and Terence A. Shimp (1990), "Effects of Involvement, Argument Strength, and Source Characteristics on Central and Peripheral Processing of Advertising," Psychology and Marketing, 7(3), 195-214.

Andrews, P. H. (1987), "Gender Differences in Persuasive Communication and Attribution of Success and Failure," Human Communication Research, 13, 372-385.

Bem, Sandra L. (1974), "The Measurement of Psychological Androgyny," Journal of Consulting and Clinical Psychology, 42 (2), 155-162.

Bem, Sandr L. (1975), "Sex Role Adaptability: One Consequence of Psychological Androgyny," Journal of Personality and Social Psychology, 31 (4), 634-643.

Bennett, Peter D., and Gilbert D. Harrell (1975), "The Role of Confidence in Understanding and Predicting Buyers’ Attitudes and Purchase Intentions," Journal of Consumer Research, 2 (September), 110-117.

Calder, Bobby J., Lynn W. Phillips, and Alice M. Tybout (1982), "The Concept of External Validity," Journal of Consumer Research, 9 (December), 240-244.

Chusmir, Leonard H., Christine S. Koberg, and Mary D. Stecher (1992), "Self-Confidence of Managers in Work and Social Situations: A Look at Gender Differences," Sex Roles, 26 (11/12), 497-512.

Cohen, Jacob and Patricia Cohen (1983), Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences, Hillsdale, NJ: Lawrence Earlbaum Associates.

Coughlin, Maureen, and P. J. O’Connor (1985), "Gender Role Portrayals in Advertising: An Individual Differences Analysis," in Advances in Consumer Research, Vol. 12, eds. Elizabeth C. Hirschman and Morris B. Holbrook, Provo, UT: Association for Consumer Research, 238-241.

Darley, William K., and Robert E. Smith (1995), "Gender Differences in Information Processing Strategies: An Empirical Test of the Selectivity Model in Advertising Response," Journal of Advertising, 24 (1), 41-56.

Dover, Philip A., and Jerry C. Olson (1977), "Dynamic Changes in an Expectancy-Value Attitude Model as a Function of Multiple Exposures to Product Information," in Contemporary Marketing Thought, eds. Barnett A. Greenberg and Danny N. Bellenger, Chicago,IL: American Marketing Association, 455-459.

Fazio, Russell H., and Mark P. Zanna (1978), "On the Predictive Validity of Attitudes: The Roles of Direct Experience and Confidence," Journal of Personality, 46 (June), 228-243.

Fazio, Russell H., and Mark P. Zanna (1981), "Direct Experience and Attitude-Behavior Consistency," Journal of Personality, 46 (June), 228-243.

Fischer, Eileen, and Stephen J. Arnold (1994), "Sex, Gender Identity, Gender Role Attitudes, and Consumer Behavior," Psychology & Marketing, 11 (2), 163-182.

Fishbein, Martin, and Icek Ajzen (1975), Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research, Reading, MA: Addison-Wesley.

Gentry, James W., and Mildred Doering (1977), "Masculinity-Femininity Related to Consumer Choice," in Contemporary Marketing Thought, eds. Barnett A. Greenberg and Danny N. Bellenger, Chicago:IL American Marketing Association, 423-427.

Gentry, James W., and Debra A. Haley (1984), "Gender Schema Theory as a Predictor of Ad Recall," in Advances in Consumer Research, Vol. 11, ed. Thomas C. Kinnear, Provo, UT: Association for Consumer Research, 259-264.

Jaffe, Lynn J. (1991), "Impact of Positioning and Sex-Role Identity on Women’s Responses to Advertising," Journal of Advertising Research, (June/July), 57-64.

Jaffe, Lynn J., and Paul D. Berger (1988), "Impact on Purchase Intent of Sex-Role Identity and Product Positioning," Psychology & Marketing, 5 (3), 259-271.

Kahneman, Daniel, and Amos Tversky (1973), "On the Psychology of Prediction," Psychological Review, 80 (4), 237-251.

Koriat, A., S. Lichtenstein, and B. Fischoff (1980), "Reasons for Confidence," Journal of Experimental Social Psychology: Human Learning and Memory, 6, 107-118.

Laczniak, Russell N. and Darrel D. Muehling (1993), "Toward a Better Understanding of the Role of Advertising Message Involvement in Ad Processing," Psychology and Marketing, 10 (4), 301-319.

Lenney, E. (1977), "Women’s Self-Confidence in Achievement Settings," Psychological Bulletin, 84, 1-13.

Maccoby, E. E, and C. N. Jacklin (1974), The Psychology of Sex Differences, Stanford, CA: Stanford University Press.

Markus, Hazel, Marie Crane, Stan Bernstein, and Michael Siladi (1982), "Self-Schemas and Gender," Journal of Personality and Social Psychology, 42 (1), 38-50.

Meyers-Levy, Joan (1989), "Gender Differences in Information Processing: A Selectivity Interpretation," in Cognitive and Affective Responses to Advertising, eds. P. Cafferata and Alice Tybout, MA: Lexington Books, 219-260.

Meyers-Levy, Joan, and Durairaj Maheswaran (1991), "Exploring Differences in Males’ and Females’ Processing Strategies," Journal of Consumer Research, 18 (June) 63-70.

Miniard, Paul W., Sunil Bhatla and Randall L. Rose (1990), "On the Formation and Relationship of Ad and Brand Attitudes: An Experimental and Causal Analysis," Journal of Marketing Research, 27 (August), 290-303.

Palan, Kay M., Charles S. Areni, and Pamela L. Kiecker (1994), "A Comparison of Gender Identity Scales in a Gift Exchange Context," working paper, Iowa State University.

Peterson, D. K., and G. F. Pitz (1988), "Confidence, Uncertainty, and the Use of Information," Journal of Experimental Social Psychology: Human Learning and Memory, 14, 85-92.

Smith, Robert E. (1993), "Integrating Information from Advertising and Trial: Processes and Effects on Consumer Response to Product Information," Journal of Marketing Research, 30 (May), 204-19.

Smith, Robert E., and William R. Swinyard (1983), "Attitude-Behavior Consistency: The Impact of Product Trial Versus Advertising," Journal of Marketing Research, 20 (August), 257-267.

Smith, Robert E., and William R. Swinyard (1988), "Cognitive Response to Advertising and Trial: Belief Strength, Belief Confidence, and Product Curiosity," Journal of Advertising, 17 (3), 3-14.

Spence, Janet T., Robert L. Helmreich, and Joy Stapp (1975), "Ratings of Self and Peers on Sex Role Attributes and Their Relation to Self-Esteem and Conceptions of Masculinity and Femininity," Journal of Personality and Social Psychology, 32 (1), 29-39.

Stern, Barbara B. (1988), "Sex-Role Self-Concept Measures and Marketing: A Research Note," Psychology & Marketing, 5 (1), 85-99.

White, M. C., G. DeSanctis, and M. D. Crino (1981), "Achievement, Self-Confidence, Personality Traits, and Leadership Ability: A Review of Literature on Sex Differences," Psychological Reports, 48, 547-569.

Wright, Peter (1975), "Factors Affecting Cognitive Resistance to Advertising," Journal of Consumer Research, 2, (June), 1-9.

Zuckerman, D. M. (1985), "Confidence and Aspirations: Self-Esteem and Self-Concepts as Predictors of Students’ Life Goals," Journal of Personality, 53, 435-560.

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