The relation between chosen role models and the self-esteem of men and women
Self-esteem has been found to correlate negatively with poor outcomes that range from eating disorders, to anxiety and depression, to gang membership (Harter, 1999; Mruk, 1999; Orvaschel, Beeferman, & Kabacoff, 1997; Shisslak, Crago, Renger, & Clark-Wagner, 1998; Wang, 1994).
One factor found to be associated with individuals' levels of self-esteem is the absence or presence of role models (Mack, Schultz, & Araki, 2002; Ochman, 1996; Pettus, 2001). In the present study we focused on the relation between the sex of an individual's chosen role models, the self-perceived difference of the role model to the self on character traits, and the individual's self-esteem.
Self-esteem can be defined simply as the way individuals feel about themselves (Steinberg, 1999). Self-esteem is also viewed as the difference between the real self and the ideal self (Pettus, 2001). If people are similar to the ideal they want to reach, they are more likely to have higher self-esteem than if they consider themselves to be far from their goal (Pettus, 2001). It has been noted that perceived inconsistencies between one's actual and ideal self-concepts are important because failure to attain an ideal may lead to serious negative outcomes (Harter, 1999). Cooper-smith (1987) suggested that the broader construct of self-esteem can be further divided into four subtypes: school/academic, home/parent, peer/social, and general self-esteem. The need to study multiple domains of self-esteem, particularly for women, has been documented by other researchers
Given that self-esteem may be considered as the difference between the real self and the ideal self, it is important to identify role models and their demographic and character traits. One characteristic of role models that may impact an individual's self-esteem is the sex of the chosen role model. Some research suggests that same-sex role models may have a more positive impact on self-esteem than other-sex role models (Ochman, 1996; Trankina, 1992). Using storybook characters to represent role models, Ochman (1996) found that third graders with same-sex role models had higher self-esteem than those with other-sex role models. Replication of these findings using actual, rather than assigned, role models would further clarify the impact that sex of chosen role models has on self-esteem. Similar findings show that the presence of same-sex role models is associated with higher self-esteem among college-aged women and predicts women's career salience and nontraditional career choices (Hackett, Esposito, & O'Halloran, 1989). Successful female role models may give younger women the self-efficacy to fulfill their own career goals.
In two studies of sex of participant and sex of role model, women were found to be more influenced by their mothers and their female friends than were men (Basow & Howe, 1980). In contrast, both men and women were influenced equally by male role models. Thus, men were more likely than women to choose same-sex role models and, although women were more likely to choose female role models than men were, they were not less likely to choose male role models. Although these findings shed light on the choice of role models by women and men, the researchers did not examine how these choices relate to self-esteem.
When an individual chooses role models who differ greatly from the self, it may adversely impact the individual's self-esteem, even though the outcome may not be overtly negative. Sometimes these role models differ greatly from the self because similar role model choices are not available. For example, people may search for others who are similar to themselves in ethnicity, age, sex, and interests in order to validate their own abilities when they make career choices but they may be unable to locate similar role models (Trankina, 1992). Research has demonstrated that ethnic minorities who are not well represented in science tend to have less confidence in their scientific abilities. The indication is that, because minorities lack role models of their ethnic group in the scientific field, their self-esteem regarding their scientific abilities is negatively impacted (Trankina, 1992).
If a discrepancy exists between the way individuals views themselves and their role models, the role model may be related negatively to self-esteem (Lockwood & Kunda, 1997, 1999). Role models may "provoke self-enhancement and inspiration when their success seems attainable, but self-deflation when it seems unattainable" (Lockwood & Kunda, 1997, p. 91). Even when individuals are similar to their chosen role models in terms of age, race, sex, and character traits, they may experience lower self-esteem if they compare themselves to their role models but feel comparatively inadequate (Tesser, 1986; Tesser & Campbell, 1983).
This study was designed to expand on the work of Basow and Howe (1980) and Ochman (1996) through an examination of the relation between the choice of other-sex or same-sex role models and levels of self-esteem. The first hypothesis was that participants would choose more same-sex than other-sex role models. The relation between the choice of male role models and women's self-esteem was of particular interest, given previous research that shows lower self-esteem among women who lack adequate female role models (e.g., Hackett et al., 1989). Therefore, the second hypothesis was that women who chose female role models would have higher levels of self-esteem than women who chose male role models.
Furthermore, the difference between character traits of the self and of the chosen role models and Coopersmith's four subtypes of self-esteem (Coopersmith, 1987) were examined. On the basis of past research findings, the third hypothesis was that perceptions by participants that they were quite different from their role models would be related to lower levels of self-esteem. Finally, as a research question, we explored the relation between the amount of self-reported influence of the role model and self-esteem, including any differences that might exist between men and women.
Participants were 120 undergraduate students who ranged in age from 18 to 33 years (M = 20.51, SD = 2.44). The sample included 60 women and 60 men. The racial distribution of the sample was 80% European American, 15% African American, 0.8% Asian, 0.8% Native American, 0.8% Hispanic, and 2.5% of participants who self-identified as "other." Participants volunteered for the study through introductory, statistical, experimental, or social psychology classes either to receive extra credit or to fulfill a research participation requirement.
The demographic form was used to collect information about participants' biological sex, age, and ethnicity.
Role Model Identification
Participants named two role models, one personally known and one famous or well known, and indicated whether each role model was male or female. Participants were asked to name the relationship between themselves and their personally known role model (e.g., my friend, my mother) and were asked to indicate what makes their well-known role model famous (e.g., author, political figure, musician). In an open-ended format, participants noted two traits that they admired about each of their chosen role models. In addition, participants rated the general influence of their role model on a scale from 1 to 10, where 1 represented the least influence and 10 represented the most influence. This method was adapted from a method described by Basow and Howe (1980) in which participants rated the amount of influence of various role models on certain decisions in their lives (e.g., career choice).
Positive Trait Ratings
Participants rated 29 adjectives (e.g., stubborn, friendly, brave), most of which were taken from Magee (2000), as to how well each adjective described themselves and each of their chosen role models, respectively. The scale ranged from 0 = strongly disagree to 5 = strongly agree. A seventh option, 6 = I don't know, was offered for the famous role model trait ratings. In the current sample, internal consistency for the total positive trait ratings scale was excellent