Gender and age patterns in emotional expression, body image, and self-esteem:

	Sign In Sign-Up

Gender and age patterns in emotional expression, body image, and self-esteem: a qualitative analysis.

Mary Polce-Lynch; Barbara J. Myers; Christopher T. Kilmartin; Renate Forssmann-Falck; Wendy Kliewer.

Sex Roles: A Journal of Research, June 1998 v38 n11-12 p1025(24)

This qualitative study used written narratives to examine gender and age patterns in body image, emotional expression, and self-esteem for a total of 209 boys and girls in the fifth, eighth, and twelfth grades. Seventy-six percent of the sample was Caucasian, 18% African-American, 5% Asian-American, and .5% Hispanic. A major finding indicates that boys restrict emotional expression from early adolescence through late adolescence, while girls increase emotional expression during the same age period. Another major finding suggests that girls in late childhood and adolescence are both more negatively and more positively influenced than boys by body image. Both boys' and girls' feelings about themselves are primarily influenced in gender-stereotyped ways.

Childhood and adolescent development involves different patterns for boys and girls in the areas of emotional expression (Nolen-Hoeksema, 1994; Pennebaker, 1995; Polce-Lynch, Myers, Kliewer, Kilmartin, 1998), body image (Abell Richards, 1996; Harter, 1989; Polce-Lynch et al., 1998), and self-esteem (Mullis, Mullis, Normandin, 1992; Polce-Lynch et al., 1998; Wade, Thompson, Tashakkori, Valente, 1989). Researchers typically measure how children and adolescents think and feel about these developmental aspects with surveys, predetermined answers, and mean scores. While such findings are valuable, much information can be lost using quantitative data alone (Trend, 1994). Additional information is needed to explore gender and age differences in how children and adolescents think and feel about themselves, their physical appearance, and the ability to express themselves. Emotional expression can be defined as the way an individual outwardly displays feelings (King Emmons, 1990). Emotional expression is considered to be an important aspect of mental and physical health (House, Landis, Umberson, 1988; King Emmons, 1990; Pennebaker, 1995). This study focused on emotional restriction, which Kilmartin (1994) defines as difficulty understanding, dealing with, and expressing emotion. An individual's ability to express thoughts and feelings can be partly reflected by the amount of information he or she expresses in verbal or written form. The ability to report feelings about the self may be a function of culturally prescribed gender roles. For example, researchers have found that adolescent boys (Polce-Lynch et al., 1998) and men (Kilmartin, 1994; Pennebaker, 1995) are more emotionally restrictive than female peers. In applied language, males' tendency to be less expressive when disclosing feelings about the self may be a function of the stereotyped "masculine" gender role that prescribes the avoidance of emotional intimacy (Shields, 1987). This disclosure pattern could inflate male self-esteem scores in that an individual who wishes to present himself in socially desirable ways may be less expressive when disclosing negative feelings about himself. More research is needed to determine which social forces influence feelings about the self, with attention given to age and gender patterns. Gender differences in feelings about the self have been reported in self-presentation research. In one study, adolescent girls' expectations of success and self-evaluations were lower compared to adolescent boys' (Gitelson, Petersen, Tobin-Richards, 1982).

Another study found college age women predicted they would have lower grade point averages (GPAs) than their male peers, although no gender difference existed in actual GPAs (Daubman, Heatherington, Ahn, 1992). Self-esteem, which refers to a person's general sense of worth and acceptance (Wylie, 1979), involves self-presentation. Similar to self-presentation findings, gender differences have also been reported in adolescent self-esteem, with boys reporting higher levels of self-esteem relative to girls (Brack, Orr, Ingersoll, 1988; Polce-Lynch, Myers, Kliewer, 1994; Wade et al., 1989) Self-esteem is considered to be a critical indicator of mental health (National Advisory Mental Health Council, 1996). Researchers have identified several influences related to self-esteem, including peer relations (Miller, 1990; Moran Eckenrode, 1991), family relations (Isberg, Hauser, Jacobson, Powers, 1989; Nielson Metha, 1994), and more recently, media influence (Polce-Lynch et al., 1998). Yet in each of these studies, influences on self-esteem were predetermined by the use of written surveys such as the Rosenberg Self-Esteem Inventory (Rosenberg, 1979).

Current literature lacks information about how children and adolescents explain what influences their self-evaluations. Other social and emotional influences related to self-esteem may have been missed due to the exclusive use of quantitative data. Body image is a major influence upon adolescent self-esteem that is consistently cited in the quantitative literature (Abell Richards, 1996; Harter, 1989; Mintz Betz, 1986; Polce-Lynch et al., 1994). In a culture preoccupied with physical appearance - as relentlessly communicated through various media - it is not surprising to find this influential relationship. In a recent quantitative study (Polce-Lynch et al., 1998), body image was a mediator, or "filter," for certain predictors of girls' self-esteem. One of these predictors was media influence. Using a self-report Likert-response format, the Media Influence Scale measured whether various media (TV, commercials, film) influence how an individual thinks and feels about his or her physical appearance. For early adolescent girls, media influence and peer relationships were related to body image, which in turn related to self-esteem. In late adolescence, media affected girls' body image, which in turn was negatively related to serf-esteem. This mediator relationship was not present for boys. However, media influence was the most significant negative predictor of self-esteem for late adolescent males.

The influential role of media with adolescent body image is receiving increasingly more attention (Kolb Albanese, 1997; Shaw Waller, 1995; Strasburger, 1995). Summary Researchers typically use quantitative research to analyze the ways children and adolescents think and feel about themselves. This study extends prior research with children and adolescents by using qualitative methods to examine gender and age differences in emotional expression, body image, and self-esteem. These methods go beyond describing "how much" of something there is, to tell us more about the essential qualities being studied (Miles Huberman, 1994). We believed that examining children's and adolescents' thoughts and feelings via an open-ended question format would allow us to examine information that might otherwise go unreported. Based on prior research, we expected boys' and girls' emotional expressiveness would differ across adolescence, with girls' being more expressive. However, since phenomenological explanations for this gender difference are lacking, we wanted to learn why it is easy or difficult for children and adolescents to express thoughts and feelings with others. This study examined emotional expression as a separate aspect of development and as a function of self-esteem, with attention given to age and gender patterns. Similarly, since a relationship has been established between body image and self-esteem, we wanted to identify the specific ways that adolescents think and talk about the influences of body image. We hypothesized that boys and girls would identify media as a prominent influence on body image. We also hypothesized that girls would report more than boys that body image is a significant influence on their self-esteem. In addition, we also expected that peers' comments about physical appearance would weigh heavily upon adolescent body image.

Finally, this study examined the broader aspect of self-appraisal: What makes children and adolescents feel good or badly about themselves? based on previous research, we expected that boys would report more positive feelings about themselves. We hypothesized that boys' more positive self-evaluation may be related to restriction of expressing negative affect. In addition to this specific hypothesis, our study also explored how it is that children and adolescents "construct" self-esteem. Moreover, what schemata do children and adolescents use to conceptualize their self-evaluations, and does this differ by gender? METHOD Participants Two hundred and nine students from private religious schools of a large Southeastern American city participated in the study. Students were enrolled in 5th grade (34 males; 41 females), 8th grade (24 males; 36 females), and 12th grade (35 males; 39 females). Of these students, 71% lived with both biological parents, 21% lived with a single mother, 4% lived with a parent and stepparent, and 1% lived with a single father; 63% lived with siblings, 3% lived with half-siblings, and 1% lived with stepsiblings. The ethnicity of the sample was primarily Caucasian (76%). African-American (18%), Asian-American (5.3%), and Hispanic (.5%) youth comprised the rest of the sample. Qualitative Questions The following questions were given to each participant: (1) "Is it easy or hard to tell others about your feelings - tell why"; (2) "How does your physical appearance or 'body image' affect how you feel about yourself"; (3) "Name some things that make you feel good about yourself"; and (4) "Name some things that make you feel bad about yourself." The purpose of these questions was to obtain written narratives regarding thoughts and feelings about the self, the influence of body image, and the ability to express affect. Procedure Researchers sent letters home and received signed parental permission to participate for students in Grades 5, 8, and 12. Students were also required to give their written informed consent.

The survey was administered by the first author in the respective schools. Participants were tested in grade groups. Total time for directions and survey completion averaged 50 minutes, with a slightly longer average time for the 5th grade classes. Quantitative data, reported elsewhere (Polce-Lynch et al, 1998), were collected at this time as well. All participants completed the survey and were guaranteed confidentiality by being assigned code numbers and by placing completed forms in a sealed envelope. A debriefing occurred after each administration and students participated in a brief discussion about the study led by the first author. Principals of each school received a copy of the research results, as did parents who elected to receive research results. In addition, each school was offered a presentation on self-esteem for parents, students, and/or teachers. Qualitative Analysis The method of analyses involved examining narrative statements and segmenting them into stems and phrases, which were then placed into categories and examined with regard to content commonalities (Tesch, 1990). This form of data reduction is common in qualitative data analysis as it allows the researcher to make contrasts and comparisons (Miles Huberman, 1994). Data were categorized by inspecting the data first, rather than identifying categories a priori. Categories more accurately reflect data when this method is used (Tesch, 1990), while Dey (1993) contends that allowing categories to emerge from data is a more accurate reflection of the content being analyzed. The process used to form categories from raw data involved reading each of the narrative statements and identifying category themes. Narrative stems and statements were then placed in thematic categories. Ratios (expressed as percentages) for each category were calculated by dividing the number of stems/sentences for a specific category theme by the total number of responses for the question. For example, 27% of 5th-grade girls' responses about what makes them feel good about themselves were categorized under "Relational Experiences." This percentage was calculated by dividing the responses in the Relational Experience category (n = 24) by the total number of responses to this question (n = 88). A similar method was used to calculate percentages when subcategory content was analyzed. When reasons were examined for why it is easy or hard to tell feelings to others, these responses were also expressed as percentages. The percentage was calculated by dividing the number of subcategory reasons by the total number of responses for the main categories of "Easy" and "Hard." For example, when explanations for why it is easy to tell feelings to others were given by 12th-grade males, 31% of the explanations involved trust. This percentage was calculated by dividing subcategory "trust" responses (n = 4) by the total number of Easy main category responses (n = 13). Responses in main categories and subcategories frequently do not equal the total number of respondents in a sample, since a respondent often gave several different stems and answers to study questions. RESULTS Emotional Expression: "Is It Easy or Hard to Tell Others About Your Feelings? Responses to this question yielded information regarding ease or difficulty with emotional expression and self-reported explanations for this judgment.

Three main categories emerged from the data: Easy, Hard, and Both (easy and hard). When written statements did not reflect a major category, an additional category of Other was used to place the few narrative themes that could not be classified into the main categories. Respondents who chose not to answer a question were tallied in the Pass category. An overview of responses to the emotional expression question is presented in Table I. As seen in this table, a general pattern of increased difficulty expressing emotions (i.e., emotional restriction), occurred for males from Grade 5 to Grade 12, while the opposite pattern was reported by females from Grade 5 to 12. This trend is best represented by their own words: a 5th-grade boy said, "It's easy, I just tell my family or people I trust"; an 8th-grade boy said, "I don't get upset"; while 12th-grade males reported, "[I] can't say it right" or "[I] have a wall up." Similar to their male peers, 5th-grade girls said it is easy to express emotions "with someone I trust like family or friends." By contrast, 8th-grade girls gave many more reasons than 8th-grade boys as to why it is easy to express emotions: "I'm open" or "It's just easy" were typical themes in 8th-grade girls' responses, as well as the notion that "It's easy to talk with friends." Responses from 12th-grade females included, "It comes naturally with people I am close to," or "It's easy to tell my boyfriend and friends - not my parents or even my psychiatrist." Twelfth-grade males reported more often that it was hard to tell their feelings. When they reported that it was easy, the reasons given involved friendships, similar to their female peers. These results indicate that males become more emotionally restrictive across adolescence, while both males and females find it easiest to talk with trustworthy friends. There was additional evidence to support a gender pattern of decreased emotional expression for males. The overall quantity of text written by females exceeded that of males in each age group. This pattern is best illustrated by comparing the total stems and sentences (raw data) as presented in Table II. Both the total output and mean responses are included in this table to represent the full "volume" of expressiveness. Although there were more girls in each sample, we believe it was valuable to include total as well as mean responses, since girls' responses typically included more volume. While the content presented in Table II involved responses to the Feel Good and Feel Bad questions, data were examined in terms of process as well as content.

We viewed the quantity of text as being related to emotional expression in that respondents were asked to disclose emotional content about themselves in written form. Specific contrasts for male groups noted an increase in total output from 5th to 12th grade with a slight "dip" in 8th grade, while the mean number of responses increased from 5th to 8th grade, leveling off in 12th grade. For girls, grade contrasts of total output noted a plateau between 5th and 8th grades, with a noticeable increase in 12th grade, while mean numbers steadily rose across adolescence for girls. Table I. Summary of Responses: "Is It Easy or Hard to Tell Your Feelings to Others? Tell Why."(a) Main Category (%) Grade and Sex Hard Easy Both Pass(b) Other 5th females (n = 42) 62(a) 17 10 12 0 5th males (n = 34) 33 26 5 21 2 8th females (n = 36) 36 39 22 0 3 8th males (n = 24) 50 17 25 4 0 12th females (n = 39) 28 44 26 0 5 12th males (n = 37) 49 35 8 14 3 a Percentages reflect respondent tallies in each main category divided by the number of respondents in that sample b This category reflects invalid responses and those who chose not to answer this question. Table II. Mean Number of Response Stems/Phrases for "Feel Good" and "Feel Bad" Questions Type of Question Feel Good(a) Feel Bad(b) Grade and Sex Total (Mean) Total (Mean) Total (Mean) 5th-grade females 86(c) (2.05)(d) 57 (1.36) 143 (3.40) 5th-grade males 53 (1.55) 31 (0.91) 84 (2.47) 8th-grade females 78 (2.17) 66 (1.83) 144 (4.00) 8th-grade males 48 (2.00) 33 (1.37) 81 (3.31) 12th-grade females 108 (2.77) 75 (1.92) 183 (4.70) 12th-grade males 77 (2.08) 45 (1.22) 122 (3.29)

a "Name some things that make you feel good about yourself.

" b "Name some things that make you feel bad about yourself."

c These numbers reflect tallies of narrativeresponses and do not include "pass" responses.

d Mean numbers reflect total responses divided by number of participants in grade and gender group.

Additional observations of responses from a process perspective revealed that males typically answered in terse and vague phrases, while females wrote detailed, full sentences. This pattern was observed-across all age groups. For example, in response to the emotional expression question, a 12th grade female said, "It's easy for me to tell others my feelings because if they did something well, I would congratulate them and tell them I feel happy for them. If someone was being mean to me I would ask them to please stop it and I feel hurt by their rudeness." This lengthy response was typical of this age and gender group. By contrast, shorter responses were the norm for early and late adolescent boys: "It is hard, I can't relate to anyone" or "easy - you have nothing to lose." Eighth-grade males were similarly terse: "Sometimes it is easy, sometimes it is hard." Explanations for Why It Is Easy or Hard to Share Feelings An important aspect of this study was to examine whether emotional expression patterns vary by age and gender - and why. The "why" responses formed subcategories as presented in Table I. When subcategory responses were examined for fifth graders, both gender differences and similarities were present. Fifth-grade boys (18%) and girls (12%) reported similarly that issues of privacy made it hard for them to tell feelings to others. For girls, additional reasons involved embarrassment (12%) and trust (21%). None of the 5th-grade boys gave embarrassment as an explanation, and only 5% mentioned the word "trust." Overall, boys' reasons were vague: "I'm not sure why" and "pass" were frequent responses. To summarize, 5th grade response patterns indicated that the majority of boys and girls find it difficult to express their feelings to others and they don't know why. For 8th-grade girls, talking with friends (50%) was the most frequently cited explanation as to why it is easy to share feelings. Other reasons ranged from "It's just easy" (14%) to "It's easy, when it's not important" (7%). Twenty-two percent of females stated that it is both easy and hard to tell feelings, noting "who it is" (63%) and "what the topic is" (38%) affected whether it is easy or hard to tell feelings. None of the girls passed on this question. For 8th-grade boys, there were no overt themes to explain why it is hard to express feelings. Rather, each boy provided his own reason for not disclosing feelings: "Emotions are illogical," "I don't know what to say," or "People will think I'm a wimp." The few responses that explained why it is easy for early adolescent boys to express their feelings included "It is easy to tell anger" and "It is just easy." Twelfth graders' explanations (Table I) for why it is easy or hard to tell feelings to others revealed that 65% of female responses cited trust as the reason for sharing feelings easily. Thirty-one percent of males' responses cited trust. The most frequent responses given by females for why it is hard to express feelings included "It is hard to talk with certain people" (36%) and "I don't like to talk about my own feelings or problems" (36%). The most frequent response given by 12th-grade males were that their personalities prevent them from expressing their emotions (18%), they are fearful or anxious of others' reactions (18%), or they can't say it right (17%), suggesting that boys have internalized gender stereotyped roles by this age. Both sexes reported difficulty putting their feelings into words (females 5%; males 16%) and that not being sure how others will react (females 12%; males 17%) makes it difficult. In general, the 12th-grade response patterns indicated that it is easier for females than males to express feelings to others due to the relationship aspect of trust. "How Does Your Physical Appearance or Body Image Affect How You Feel About Yourself?" Although body image emerged as a main category for responses to good and bad feelings about the self, we used this question to directly ascertain how body image affects the self. There was some repetition of results, which tended to validate grade and gender patterns. Responses to this body image question ranged from an 8th-grade boy who stated, "It doesn't [affect me]," to a 5th-grade girl who said, "It makes me feel like I shouldn't be on this earth." The four main categories that emerged from the data were Positive Effect, Negative Effect, No Effect and Somewhat (affects me). As before, the category of Other included themes that did not fit into a main category, while the category of Pass represented those who chose not to answer. Table III presents a summary of the qualitative data as organized by category. As seen in Table III, there were large age and gender differences in response to how body image affects feelings about the self. Eighth-grade girls most often reported that body image has a negative effect, while their male peers reported most often that body image has no effect upon them. Girls' comments in early adolescence often focused on body parts and the relationship between appearance and mood, i.e., "I like the way I look, I just wish my butt wasn't quite so big and my chest wasn't quite so flat," or "If I feel I look bad, I'll be in a pretty bad mood." Fifth- and twelfth-grade boys' and girls' responses were evenly distributed with regard to the negative effect of body image. However, those in the older group reported more often that body image has a negative effect upon them. [TABULAR DATA FOR TABLE III OMITTED] Influences on Self-Esteem "Name some things that make you feel good about yourself." Four general categories emerged from the data in response to this question for males and females in all grades: (1) relational experiences with family, friends, and other persons in one's life; (2) self-generated behaviors such as grades, sports, and talents; (3) body image issues such as looks, weight, and clothes; and (4) personality or trait characteristics such as likeable, smart, kind, etc. Categories of Other and Pass were created for purposes previously described. A summary of gender and grade comparisons for the "feel good" question is presented in Table IV. Within the Relational Experiences category, "family" was the most frequent response for each age and gender group, followed closely by "friends." Examples of comments included, "having family and friends that love me" (8th-grade boy), or "my family, my friends" (5th-grade girl). Table IV also shows how responses with a "personality" theme tended to differ by grade and gender. Younger boys reported more often than twelfth-grade males that their personalities have a positive influence on their self-evaluations. Eighth-grade boys included mention of being talented, intelligent, or excelling in math and music, while 12th-grade males reported these same traits or abilities, but less frequently. This age pattern was reversed for girls in that the adolescent girls (8th and 12th graders) reported more often than younger girls (5th graders) that personality traits are a source of positive feelings about the self. It is also worth noting that males in all grades frequently reported "humorous/funny," while 5th- and 8th-grade girls listed "smart" as the most frequent response. Twelfth-grade [TABULAR DATA FOR TABLE IV OMITTED] females equally reported both "funny" and "smart." The second most frequent response for males in Grades 5 and 12 was intelligence. Gender differences emerged with themes of body image and how it affects feelings toward the self without direct prompting in the question. In other words, when asked in an open-ended question, body image surfaced as a main theme. These results replicated previous body image results, suggesting that body image has different meanings for boys than for girls. Females, more often than males in all grade levels, cited overall looks and specific body parts (i.e., hair, face, smile) as a source of feeling good about the self. Girls in late childhood and early adolescence consistently reported more than boys that their physical appearance made them feel good about themselves, but in late adolescence this gender difference leveled off. In other words, males and females in 12th grade reported that they all felt about the same regarding the positive influence of physical appearance.

"Name some things that make you feel badly about yourself." There was a noticeable gender difference between data sets for the "feel bad" question and the "feel good" question for each grade. Males and females in all age groups named fewer things that made them feel bad about themselves compared with things that influenced them positively. Males restricted their expression more so than females. Grade and gender response patterns for this question are summarized by category in Table V. (Also, see Table II for additional comparisons.) As seen in Table V, body image responses were similar for boys and girls in 5th and 12th grades. Eighth-grade females reported more frequently than any other age group that body image was a source of bad feelings about the self. For males and females in all grades, the most frequent response within the relational experiences category involved hurt and rejection [TABULAR DATA FOR TABLE V OMITTED] by peers and family. Fifth grade females often stated specifically that "people making fun of me" made them feel badly. With age, reasons became more sophisticated as evidenced in these 12th-grade responses: "betrayal by friends" or "when my parents compare me to super students." Males and females across age groups also reported similarly within the self-generated category, citing "grades" and "things I do wrong/doing wrong" most often. As indicated in Table V, none of the boys and girls in late childhood and early adolescence reported personality characteristics as being a source of negative feelings about the self. However, late adolescent females did report personality traits such as being "impatient," "unassertive," "stubborn," "irritable," etc., as being a source of negative feelings about the self, while their male counterparts did not. Another age and gender difference involved body image as a source of bad feelings about the self. Within the body image category, one third of 5th-grade girls' responses indicated that being "too heavy" made them feel badly. The other two-thirds focused on overall concerns of being "ugly." Twelfth-grade females focused more on "body parts" (e.g., legs, thighs) and "weight" than on overall appearance as having a negative effect on feelings about the self. This was also true for 12th-grade males, but they gave examples of "hair" and "skin color" rather than body parts. It is important to note that the racial composition of this grade was homogeneous and Caucasian.

DISCUSSION

This study investigated how children and adolescents think and feel about themselves, their emotional expression, and their body images by, eliciting "their own words." We used open-ended questions to capture information that might otherwise be obscured by a close-ended, Likert-type format. We were interested in identifying developmental gender similarities and differences. Emotional Expression As expected, the most accentuated finding in this study involved gender and age differences in emotional expression. The increased emotional restriction in adolescent males and increased emotional expression in adolescent females suggested that boys and girls follow socially prescribed "display rules" that are gender specific. It is worth noting here that quantitative data from the same sample revealed a similar gender and age pattern, with boys becoming significantly more restrictive across adolescence (Polce-Lynch et al., 1998).

This contrast in response style seemed to support the presence of restricted emotional expression in boys, in keeping with the stereotyped masculine display rule, which implies that masculinity is reduced when boys talk about feelings. Or more pointedly, that they are "sissies" and "wimps" when they break this gender "rule." Similarly, increased emotional expression for adolescent gifts may reflect compliance with the culturally prescribed feminine display rule that encourages gifts to do "gift talk." When boys and girls follow these display rules, it may be more difficult to obtain true thoughts and feelings about emotionally charged content from adolescent boys, while it appears easier to elicit accurate information from adolescent gifts.

Although it can also be argued that girls' self-presentation with certain disclosures may not be accurate either (Daubman et al., 1992). Since there was a "dip" in emotional expression for 8th-grade boys and a "plateau" for girls, it may be that early adolescence is a time when both girls and boys are generally less expressive. Research from the developmental literature on transitions related to puberty suggests that not only is this a difficult time with regard to social and emotional development, but that traditional gender roles may be intensified in early adolescence (Hill Lynch, 1983; Steinberg, 1981). Relatedly, it is of interest that girls typically score lower than boys on measures of mental health such as body image, peer relations, and self-esteem during early adolescence (cf. Polce-Lynch et al., 1994; Polce-Lynch et al., 1998).

This early adolescent pattern may be explained by cognitive developmental differences (Maccoby Jacklin, 1974; Steinberg Hill, 1978), which call attention to girls' greater facility with verbal tasks, i.e., reporting thoughts and feelings is easier for them. Gifts probably have more "practice" expressing themselves via talking on the phone, writing notes, letters, and diaries at this age. Or, it could be that gender differences in reporting reflect gender stereotyped self-presentation styles (cf. Daubman et al., 1992; Gitelson et al., 1982) and other elements of social construction. This dichotomy reflects an ongoing debate between the essentialist and the constructivist perspectives. However, there is strong empirical support for the view that differences between individuals are greater than gender differences, especially with regard to social behaviors (Basow, 1992; Lott, 1994). This study did not identify outcomes associated with emotional restriction.

However, if thought and emotion are intertwined aspects of the same process, as theorists currently suggest (Haviland et al., 1994; Powers, Welsch, Wright, 1994), increased emotional restriction in males seems to be a risk factor for mental illness and relationship problems. This hypothesis is supported by Hendryx, Haviland, and Shaw (1991) and Kilmartin (1994), among many others. Some theorists and researchers suggest that boys' use of blunting, distraction, or denial of negative feelings may actually prevent depression and other affective disorders (Kliewer, 1991; Nolen-Hoeksema, 1994). Yet, one could also argue that such distraction is unhealthy repression or dissociation of affect. In other words, distraction-type responses "bury" a problem rather than "working it through." When males do not experience feelings and when they do not cope directly with affect, it seems plausible to suggest that the affective energy must go somewhere (Lynch Kilmartin, 1998; Real, 1997). This mind-body link is currently receiving empirical attention. Behavioral psychology and behavioral medicine researchers are establishing both a theoretical basis and strong empirical evidence for a correlational impact between social relationships and health (House et al., 1988; Pennebaker, 1995; Wickramasekera, 1993).

Later in life, males' emotional restriction may render them more susceptible to health problems because their bodies unconsciously do the "emotional" work for them - in the form of medical problems such as ulcers, high blood pressure, heart problems. Females who are emotionally restrictive can also be negatively affected this same way. The point made here is that, generally speaking, cultural display rules restrict men's emotional expression more so than that of women. Thus, the etiology of men's mental and physical health risks may be somewhat different than women's due to emotional restriction. If restricted emotionality is one risk factor for mental and physical health, then interventions must begin early in the socialization process for males. Boys in late childhood were not much different from girls, but by early adolescence, boys were well entrenched in the gender role stereotype of restricting emotions. Many statements from late adolescent males suggest that they are very confused and struggling when it comes to "emotional stuff" and relationships. Males know too well that outward display of emotion is "out role" behavior in our culture, and that it makes people feel uncomfortable and can result in peer rejection (Kilmartin, 1994). So, even if adolescent males do express their feelings, they may not get the social support they need. By contrast, females who express emotions may be less at risk for mental and physical health problems because of the social support they elicit from listeners in our culture. It is clear that we need to find ways to help boys retain emotional expressiveness. According to Kilmartin (1994), the reasons that boys and men do not readily express their emotions are quite succinct: personality, punishment, and skill deficit. If this is true, then the remedies are equally clear. We must help boys and men develop the vocabulary to "live in" the emotional world, and we need to reward - rather than punish - appropriate and healthy emotional expressiveness in school, family, and professional relationships. Body Image Since body image, or physical appearance, has been identified as a significant predictor of adolescent self-esteem in previous quantitative research (Abell Richards, 1996; Harter, 1989; Polce-Lynch et al., 1994; Polce-Lynch et al., 1998), we were interested in learning more about underlying thoughts and feelings about body image.

As expected, adolescent girls reported that they are both positively and negatively affected by body image more than boys. This finding is consistent with Wood's (1997) conclusion that, although both men and women are affected by cultural norms of physical attractiveness, and while men may dislike their physical features, their judgments about these attitudes seldom affect overall self-esteem the way they do for women. We had hypothesized that media influence would be reported as an influence on body image based on prior research (cf. Kolb Albanese, 1997; Shaw Waller, 1995) Yet, only one girl in the entire study reported "media" as an influence on body image. This omission was even more surprising since quantitative data from the same sample revealed that the most significant negative predictor of self-esteem for 5th-grade girls and 12th-grade boys was how media influenced their physical appearance (Polce-Lynch et al., 1998). In addition, media influence was significantly correlated with self-esteem for all groups except early adolescent boys. The total absence of boys' and girls' reports that media is an influence on body image (when asked in an open-ended format) is stunning. While other explanations may exist, we strongly believe that this "nonfinding" reflects cognitive developmental levels (i.e., newly emerging abstract thought processes) as well as levels in the development of conscious awareness. Consciousness plays a critical role in raising awareness regarding gender stereotypes and media's influence. In other words, children and adolescents are vulnerable to mass media because they do not think about the messages communicated to them. Therefore, unless children and adolescents are taught the importance of - and how to - deconstruct mass media (TV/MTV, commercials, advertisements, etc.), this aspect of culture will largely remain an unconscious influence on body image, allowing media's influence to be all the more insidious. Indeed, Shaw and Waller (1995) have reviewed evidence that links media's impact upon body image, and they suggest educators and parents should encourage adolescents to question the acceptance of "ideal" media images. In a discussion of body image and media influence, it is important to call attention to the body image literature that includes evidence that beauty issues tend to affect blacks and whites differently. More specifically, one study indicated that African-American males' self-evaluations were more positively associated with lighter skin color than females' (Wade, 1996). In addition, previous studies have indicated that African-American women are better at resisting cultural messages of physical attractiveness (Thomas, 1989; Thomas James, 1988). This may be related to Root's (1990) findings that women's weight is judged less negatively by African American men than by Caucasians. Another explanation could be that African American women who identify strongly with an ethnic heritage that values full-figured bodies are grounded in an identity that helps them resist white culture's emphasis on thinness. It is also important to add that Lott's (1994) research review suggests that middle and upper class minority women may be becoming more susceptible to unhealthy views of weight and body image.

This brings to light the role of class in beauty and body image issues. While racial comparisons were not made due to the small proportion of racial minority participants in this study, future research should expand upon previous findings by identifying gender and developmental patterns in body image for both whites and minorities, with particular attention given to the role of class. Some interesting age and gender patterns emerged regarding the positive and negative role of body image in children's and adolescents' experiences. One salient theme for adolescent boys and girls was the relationship between appearance and mood, i.e., "When I look good I feel good" and "When I look bad, I feel bad." These responses indicated that when adolescents have a "bad hair day," this outward aspect of physical appearance generalizes inward. This process might be expected in early adolescence when thoughts and feelings can be telescopic. But, surprisingly, late adolescents were still strongly affected in this same way. These findings were consistent with previous research that has linked the role of media to promotion of certain standards of attractiveness for women (Silverstein et al., 1986; Wood, 1997). Mass media's telescopic focus on "air-brushed" body parts and overall appearances, probably contributes to the connection between adolescents' perceptions of physical appearance and their inner feelings. Specific response patterns revealed that 8th-grade girls were the most vocal of any group about the negative effects of body image. Their remarks included themes of heightened self-awareness in a self-critical way. A representative example was, "I feel like I stick out in a bunch of crowded normal people." By contrast, boys in late childhood and late adolescents reported fewer negative effects. It appeared that boys and girls are most similar in late childhood and then again in late adolescence. As previously cited, it may be that early adolescence, coinciding with the onset of puberty, represents a time of gender role intensification - with "masculinity" and "femininity" becoming accentuated polarities. A recent longitudinal study indicated that gender roles are more stereotyped in early adolescence than in late adolescence (Alfieri, Ruble, Higgins, 1996). Fifth-grade boys reported more often than female peers that body image has a positive effect for them.

In late adolescence, however, females reported more frequently that body image has a positive effect on them. In other words, younger boys and older female teens generally report positive experiences related to their physical appearances, while older males report fewer - or more negative - experiences related to their physical experiences. This finding is consistent with the pattern of reporting differences identified in this study, i.e., that as boys get older they restrict the expression of affectively charged information about themselves. By contrast, we know that physical appearance is clearly on girls' minds. The combination of cultural display rules and objectification of the female body denies boys the help they may need with regard to body image, while it simultaneously prepares adolescent girls for the process of internalizing the object of gaze. Frederickson and Roberts (1997) present an interesting theory of this objectification process for girls and women. Influences on Self-Esteem: Good and Bad Feelings About the Self This aspect of the study was more exploratory in that we were interested in learning what boys and girls identified as influences upon their self-esteem. based on previous quantitative research, we expected that girls would report more often than boys that body image and relationships influenced feelings about the self. When comparing the overall themes of what makes boys and girls feel good about themselves, we learned that across age groups, both boys and girls reported relational experiences and self-generated behaviors to be major influences.

However, girls reported more frequently than boys that relational experiences influence them. This finding suggests that experiences with others such as family and friends, and behaviors such as good schoolwork and sports, create positive self-regard. Thus, relationships and activities are important in promoting positive feelings and enhancing self-esteem. A specific developmental gender pattern regarding positive feelings about the self involved body image. Although already discussed as a separate issue, body image is highlighted again here because children and adolescents reported it to be an influence upon self-evaluations. Overall, body image was important in positive and negative self-appraisals for girls, while for males in all age groups, body image was reported, albeit less often, as a source of esteem. These results were consistent with the separate body image narrative themes, suggesting that these gender and age patterns may reflect a reified trend. This trend seems to parallel the portrayal of beauty and the female body in television, film, and advertisements.

However, now it appears that males are also becoming the object of gaze, as evidenced in recent advertisement research (Kolb Albanese, 1997). The important message that both boys and girls (as young as 5th grade) are giving is that their physical appearance significantly influences them, both positively and negatively. Results from this study substantiate Pipher's (1994, 1996) contentions that children and adolescents must learn to value and evaluate themselves from a multidimensional perspective, rather than by their physical appearances alone. The gender difference in how "personality" traits influenced self-esteem may represent the effect of gender stereotypes. Females valued intelligence as a primary source of positive feelings about the self across all age groups. Following the literature on gender stereotypes, analytical intelligence is typically associated with masculinity (Alfieri et al., 1996; Bem, 1974), so the girls' responses may, in part, be evidence of change in this stereotype. The results also suggested that boys in late childhood tend to internalize positive aspects of themselves more than late adolescent males, while the reverse appears to be true for females. Indeed, this suggests that late adolescent females may tend to internalize negative aspects of themselves more than younger girls or male age peers. This age and gender difference probably reflects culturally normed "bragging rights" (cf. Gitelson et al., 1982). In American culture, boys are given more permission to boast about themselves, while girls are socially scripted to be more self-effacing. This behavioral influence is best substantiated by noting prevalent masculine stereotypes: "competitive," "dominant," and "self-confident" (Bem, 1974; Street, Kimmel, Kromrey, 1995).

By contrast, our culture's traditional feminine stereotypes involve "yielding," "dependent," and "sensitive" character traits (Bem, 1974; Street et al., 1995). In short, it appears that boys are better primed than girls when it comes to the process of internalizing positive aspects of the self. Overall, terse answers to the "feel bad" questions indicated that it was difficult for boys and girls to identify what makes them feel badly about themselves. Despite overall similarities among the groups, there were pronounced age and gender differences. For example, 5th-grade girls reported most frequently that relational experiences made them feel badly. Questions about whether girls in late childhood have adequate coping skills to externalize relationship problems, or whether these girls actually experience and report more negative peer relationships than do boys, were left unanswered by our study.

However, Crick and Grotpeter (1995) suggest that girls at this age experience a type of relational aggression that boys do not. Another gender difference in response to the "feel bad" question involved personality type responses. Late adolescent girls attributed "feel bad" responses to their own personality characteristics more often than their male peers and more often than any other age or gender group. At the same time, these young women gave an equal percentage of personality responses to the "feel good" questions, suggesting that older adolescent girls may be better able to integrate both positive and negative aspects of themselves. Indeed, this finding is consistent with previous research that has identified girls' earlier maturation than boys' in ego development and personality traits (cf. Cohn, 1991). In contrast, early adolescent boys frequently reported that their behaviors make them feel badly about themselves. Although they were the least emotionally expressive group, they were able to disclose this negative influence. We wondered whether this was another expression of culturally prescribed gender roles. Boys may be more involved with and more sensitized to activities with judgment-type outcomes. This could be considered the flip side of "bragging rights." In other words, a boy's identity and self-worth may become associated with not making the touchdown, the field goal, or the traveling team. This explanation would be consistent with Shields' (1987) research, suggesting that it is considered okay for boys to express negative emotions within certain domains. Thus, they may be more likely to report negative feelings regarding their behaviors. Limitations and Future Research Perhaps the main limitation of this study involves the use of self-report data. The assumption that individuals can accurately provide reasons or explanations for their behaviors has been challenged in various forums (cf. Nisbett Wilson, 1979). Yet, when seeking the actual thoughts and feelings of an individual, direct information can only be given by the individual.

We believe the actual words of children and adolescents in this study reflected the schemata they use to conceptualize themselves. It is worth noting here that age and gender patterns in results of the companion quantitative study were similar for this sample (Polce-Lynch et al., 1998). Another possible limitation involves the fact that participants attended private schools. While it could be argued that the sample is still exposed to the larger culture, replicating these questions with a public school sample would provide additional insight into emotional expression, body image, and influences on the self. In addition, some researchers have noted that the issue of body image is more salient in private homogeneous schools (McCaulay, Mintz, Glenn, 1988), so this sample may actually clarify the role of adolescent body image, rather than being biased. Since this study used a cross-sectional design, future research should include replication with a longitudinal method to detect age changes. It would also be interesting to extend this study into adulthood in order to compare age differences or age changes in body image, self-esteem, and emotional expression.

Given the large gender difference in emotional expression, it seems important to follow up and determine how emotional expression influences various psychological outcomes and life situations. Late adolescent males seemed to be struggling with their inability to express emotions, especially in the context of intimacy. Future research is needed to identify interventions that will increase males' healthy emotional expression. And, based on empirical indicators, interventions must begin early since the socialization process that restricts boys' emotional expression appears to be entrenched by early adolescence. Future research directions should also focus on the strong influence of body image and physical appearance upon girls' self-evaluations. Prevention and intervention efforts from parents and professionals are clearly needed to lessen the significant negative impact of body image. These efforts must begin with girls prior to 5th-grade, and should continue throughout adolescence. More specifically, girls, as well as boys, would benefit from learning how to deconstruct and critically evaluate media's messages about physical appearance. But before this deconstruction process can even begin, we must do the more basic work of raising consciousness levels about the basic ways (unhealthy) gender stereotypes are perpetuated in the media. Conclusion The most important contribution of this study may be that we listened to what children and adolescents had to say about themselves - and that they spoke. Despite recent efforts from the psychology of men and almost three decades of work from the psychology of women, it appears that gender stereotypes are alive and well. Returning to the research questions of how and why body image and emotional expression develop differently for boys and girls, one answer appears to be that mainstream American culture makes its way into our youth's development. In other words, gender roles and stereotypes dictate emotional display rules for boys and girls, and physical appearance is far more important than it should be to girls.

Fortunately, the youth in this study shared some of their self-esteem antidotes: relationships matter, and so does being productive and being involved in healthy activities.

REFERENCES

Abell, S. C., Richards, M. H. (1996). The relationship between body shape satisfaction and self-esteem: An investigation of gender and class differences. Journal of Youth and Adolescence, 25, 691-703. Alfieri, T., Ruble, D. N., Higgins, E. T. (1996).

Gender stereotypes during adolescence: Developmental changes and the transition to junior high school. Developmental Review, 32, 1129-1137. Basow, S. (1992).

Gender: Stereotypes and roles (3rd ed.) Monterey, CA: Brooks/Cole. Bem, S. L. (1974).

The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 155-162. Brack, C., Orr, D., Ingersoll, G. (1988).

Pubertal maturation and adolescent self-esteem. Journal of Adolescent Health Care, 9, 280-285. Cohn, L. D. (1991).

Sex differences in the course of personality development: A meta-analysis. Psychological Bulletin, 109, 252-266. Crick, N. R., Grotpeter, J. K. (1995).

Relational aggression, gender, and social-psychological adjustment. Child Development, 66, 710-722. Daubman, K. A., Heatherington, L., Ahn, A. (1992)

Gender and the self-presentation of academic achievement. Sex Roles 27, 187-196. Dey, I. (1993).

Qualitative data analysis: A user friendly guide for social scientists. New York: Routledge. Frederickson, B. L., Roberts, T. (1997).

Objectification theory: Toward understanding women's lived experiences and mental health risks. Psychology of Women Quarterly, 21, 173-206. Gitelson, I. B., Petersen, A. C., Tobin-Richards, M. H. (1982).

Adolescents' expectancies of success, self-evaluations, and attributions about performance on spatial and verbal tasks. Sex Roles, 8, 411-419. Harter, S. (1989).

Adolescent self and identity development. Unpublished manuscript, University of Denver. Haviland, J. M., Davidson, R. B, Ruetsch, C., Gebelt, J. L., Lancelot, C. (1994).

The place of emotion in identity. Journal of Research on Adolescence, 4, 503-518. Hendryx, M. S., Haviland, M. G., Shaw, D. G. (1991).

Dimensions of alexithymia and their relationships to anxiety and depression. Journal of Personality Assessment, 56, 227-237. Hill, J., Lynch, M. (1983).

The intensification of gender-related role expectations during early adolescence. In J. Brooks-Gunn A. Petersen (Eds.),

Girls at puberty. New York: Plenum. House, J. S., Landis, K. R., Umberson, D. (1988). Social relationships and health. Science, 241, 540-545. Isberg, R., Hauser, S., Jacobson, A., Powers, S. (1989).

Parental contexts of adolescent self-esteem: A developmental perspective. Journal of Youth and Adolescence, 18, 1-23. Kilmartin, C. T. (1994).

The masculine self. New York: Macmillan. King, L., Emmons, R. (1990).

Conflict over emotional expression: Psychological and physical correlates. Journal of Personality and Social Psychology, 58, 864-877. Kliewer, W. (1991).

Coping in middle childhood: Relations to competence, Type A behavior, monitoring, blunting, and locus of control. Developmental Psychology, 27, 689-697. Kolb, R. H., Albanese, P. J. (1997).

The functional integration of sole-media images into magazine advertisement. Sex Roles, 36, 813-836. Lott, B. (1994).

Women's lives: Themes and variations in gender learning (2nd ed.). Pacific Grove, CA: Brooks/Cole. Lynch, J. R., Kilmartin, C. T. (1998).

The pain behind the mask. Masculine depression: Origins, consequences, and remedies. Binghamton, NY: Haworth Press. Maccoby, E., Jacklin, N. (1974). T

he psychology of sex differences. Stanford, CA: Stanford University Press. McCaulay, M., Mintz, L., Glenn, A. A. (1988).

Body image, self-esteem, and depression-proneness: Closing the gap. Sex Roles, 18, 381-391. Miles, M. B., Huberman, A.M. (1994).

Qualitative data analysis. Thousand Oaks, CA: Sage. Miller, K. (1990). Adolescents' same-sex and opposite-sex peer relations: Sex differences in popularity, perceived social competence, and social cognitive skills. Journal of Adolescent Research, 5, 222-241. Mintz, L. B., Betz, N. E. (1986).

Sex differences in the nature, realism, and correlates of body image. Sex Roles, 15, 185-195. Moran, P. B., Eckenrode, J. (1991).

Gender differences in the costs and benefits of peer relationships during adolescence. Journal of Adolescent Research, 6, 396-409. Mullis, A., Mullis, R., Normandin, D. (1992).

Cross-sectional and longitudinal comparisons of adolescent self-esteem. Adolescence, 27, 51-61. National Advisory Mental Health Council. (1996).

Basic behavioral science research for mental health: Vulnerability and resilience. American Psychologist, 51, 22-28. Nielson, D., Metha, A. (1994).

Parental behavior and adolescent self-esteem in clinical and non-clinical samples. Adolescence, 29, 525-542. Nisbett, R. E., Wilson, T. D. (1979).

Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84, 231-259. Nolen-Hoeksema, S. (1994).

An interactive model for the emergence of gender differences in adolescence. Journal of Research on Adolescence, 4, 519-534. Pennebaker, J. W. (1995).

Emotion, disclosure, and health: An overview. In J. W. Pennebaker (Ed.), Emotion, disclosure and health. Washington: American Psychological Association. Pipher, M. (1994).

Reviving: Saving the selves of adolescent girls. New York: Ballantine Books. Pipher, M. (1996). In the shelter of each other. New York: Ballantine Books. Polce-Lynch, M., Kliewer, W., Myers, B. J. (1994).

Gender differences in early adolescent self-esteem: The mediating role of body image. Unpublished manuscript, Virginia Commonwealth University, Richmond, VA. Polce-Lynch, M., Myers, B. J., Kliewer, W. Kilmartin, C. T. (1998).

How contemporary influences and gender relate to adolescent self-esteem. Manuscript submitted for publication. Powers, S. R., Welsh, D. P., Wright, V. (1994).

Adolescents' affective experience of family behaviors: The role of subjective understanding. Journal of Research on Adolescence, 4, 585-600. Real, T. (1997).

I don't want to talk about it: Overcoming the secret legacy of male depression. New York: Scribner. Root, M.P. (1990). Disordered eating in women of color. Sex Roles, 22, 525-536. Rosenberg, M. (1979).

Conceiving the self. New York: Basic Books. Shaw, J., Waller, G. (1995).

The media's impact on body image: Implications for prevention and treatment. [Special Issue: Body Experience]. Eating Disorders: The Journal of Treatment and Prevention, 3 115-123. Shields, S. (1987).

Women, men, and the dilemma of emotion. In P. Shaver C. Hendrick (Eds.), Sex and gender. Newbury Park, CA: Sage. Silverstein, B., Perdue, L., Peterson, E., Kelly, E. (1986).

The role of the mass media in promoting a thin standard of body attractiveness for women. Sex Roles, 14, 519-523. Steinberg, L. (1981).

Transformations in family relations at puberty. Developmental Psychology, 17, 833-840. Steinberg, L., Hill, J. (1978).

Patterns of family interaction as a function of age, onset of puberty, and formal thinking. Developmental Psychology, 14, 683-694. Strasburger, V. C. (1995).

Adolescence and the media: Medical and psychological impact. New York: Ballantine Books. Street, S., Kimmel, E. B., Kromrey, J. D. (1995).

Revisiting university student gender role perceptions. Sex Roles, 33, 183-201. Tesch, R. (1990). Qualitative research: Analysis types and software tools. New York: Palmer Press. Thomas, V. G. (1989).

Body image satisfaction among black women. Journal of Social Psychology, 129, 107-112. Thomas, V. G., James, M.D. (1988). Body image, dieting tendencies, and sex-role traits in urban black women. Sex Roles, 18, 523-529. Trend, M. G. (1994).

Notes from the field: On the coordinated use of quantitative and qualitative data. In S. H. Cohen H. W. Reese (Eds), Lifespan developmental psychology: Methodological considerations, Hillside, NJ: Lawrence Erlbaum Associates. Wade, T. J. (1996).

The relationship between skin color and self-perceived global, physical, and sexual attractiveness, and self-esteem for African Americans. Journal of Black Psychology, 22, 358-373. Wade, T. J., Thompson, V., Tashakkori, A., Valente, E. (1989).

A longitudinal analysis of sex by race differences in predictors of adolescent self-esteem. Personality and Individual Differences, 10, 717-729. Wickramasekera, I. (1993).

Assessment and treatment of somatization disorders: The high risk model of threat perception. In S. Lynn, J. Rhuye, I. Kirsh (Eds.), Handbook of clinical hypnosis. Washington, DC: American Psychological Association. Wood, J. T. (1997).

Gendered lives: Communication, gender and culture. Belmont, CA: Wadsworth. Wylie, R. (1979). The self-concept, (Vol. 2). Lincoln: University of Nebraska Press.

articles references index