The impact of sex and gender role self-perception on affective reactions to different types of film
Mary Beth Oliver
From the moment of birth, a child’s world is organized along gender-related lines. The blue- vs. pink-colored blankets used to swaddle infants, the nursery rhymes that describe what boys and girls are made of, and the sex-based segregation in elementary classrooms are but a few of the hundreds of behaviors that underline the importance of gender as a classification scheme. Given the pervasiveness of gender differentiation, it comes as no surprise that the media also participate in dividing up the world in ways that both reflect and perpetuate stereotypes of gender (see Davis, 1990; Frueh & McGhee, 1975; Gerbner, Gross, Morgan, & Signorielli, 1986; Thompson & Zerbinos, 1995). Although a substantial amount of media research has explored the content of gender-stereotyped portrayals and their effects on viewers, it is important that researchers also consider the role of viewers’ responses to and enjoyment of such content because (1) entertainment fare is often targeted specifically to male vs. female audiences; and (2) differential viewing of media entertainment may serve to exacerbate sex role stereotyping and behavior differences. Consequently, the purpose of the present study is to extend previous research on gender and the media by exploring the roles that both biological sex and psychological gender roles (communal and agentic traits) play in viewers’ affective responses to media entertainment.
Responses to Media Entertainment
The appeal of media entertainment is easily demonstrated in terms of the sheer amount of time it consumes in the lives of most individuals. Recent appraisals of leisure activities estimate that Americans spend almost half of their free time, more than 20 hours per week, devoted to media consumption including television, magazines, books, radio, and film (Jeffres, 1994). Although it is clear that media offerings are enjoyed by many people, it is much more difficult to assess the specific types of media offerings that are enjoyed and the specific viewer characteristics that predict responses. Viewers’ responses to media entertainment likely reflect a complex combination of personality traits, social situations, and media-content characteristics. Given the seemingly infinite number of factors affecting viewers’ responses, it is clearly a very difficult task to explain why one viewer screams in terror during a horror film while the next shrugs in indifference, or why one viewer sobs uncontrollably during a tear-jerker while another yawns from boredom.
Research on responses to entertainment has explored a variety of personality, social, and content factors thought to be predictive of viewers’ reactions. For example, Weaver (1991a; also see Weaver, Brosius, and Mundorf, 1993) studied personality characteristics such as neuroticism, psychoticism, and extraversion, and reported that these variables successfully predicted many media preferences. Zillmann, Weaver, Mundorf, and Aust (1986) explored social and gender role variables and reported that the emotional responses of one’s opposite-sex viewing partner have strong effects on one’s own responses to frightening films. Similarly, Mundorf, Weaver, and Zillmann (1989) found strong effects on the estimation of an opposite-sex target’s reactions to horror films as a function of both the raters’ sex and their gender role characteristics (for a review of this literature, see Zillmann & Weaver, 1996). Finally, Tamborini, Stiff, and Zillmann (1987) explored the effects of content characteristics on viewers’ responses to graphic horror films, and reported that the gender of the characters affected viewers’ preferences.
Despite the fact that research has employed divergent approaches and has explored numerous variables, the one variable that has likely been examined most frequently is the sex of the viewer. Examples of sex differences in responses to entertainment are consistent and numerous. Richards and Sheridan (1987), for instance, examined the appeal of different movie genres using a sample of over 550 male and female movie goers. They found that women gave the highest priority to romance, history, and love story themes while men preferred presentations focusing on action, crime, and sex. Interestingly, the respondents’ rationalization for their film preferences provided via interviews highlighted the importance of gender considerations in many viewers’ decisions. Generally, men responded that it was very “natural” for them to prefer action-packed films over romantic ones simply because they were of the “masculine” gender. In contrast, women responded that they preferred love stories because such films touched and moved their hearts, thus bringing out their “feminine” traits. Similarly, other studies have reported that females evidence greater fear and less enjoyment of frightening films than do males (Cantor & Reilly, 1982; Sparks, 1986; Tamborini & Stiff, 1987). On the other hand, males report less involvement, interest, emotional responsiveness, and enjoyment of sad films or tear-jerkers than do females (see Oliver, 1993). The fact that sex plays such a robust and recurrent role in viewers’ responses to media entertainment makes it a variable worthy of further exploration.
Sex Differences in Emotion and Empathy
Stereotypes of sex often include the belief that females are generally more emotional than are males (Antill, 1987; Birnbaum, Nosanchuk, & Croll, 1980; Broverman, Vogel, Broverman, Clarkson, & Rosenkrantz, 1972; Frieze, Parsons, Johnson, Ruble, & Zellman, 1978). Although serf-reports of emotion are frequently consistent with these beliefs, the type of emotion in question plays an important role in the direction of the difference. That is, females tend to report more intense feelings of sadness and fear (and sometimes joy) than do males, while males tend to report more intense feelings of anger (Allen & Haccoun, 1976; Lombardo, Cretser, Lombardo, & Mathis, 1983; Shields, 1987; Staley & O’Donnell, 1984). In a recent literature review of sex differences and emotion, Brody and Hall (1993) summarized previous research as suggesting that females are more likely than males to express emotions associated with affiliation and social bonding (e.g., warmth, pity, and sadness), whereas males are more likely than females to express emotions associated with differentiation and competition (e.g., anger and pride; see also Brody, 1993).
Consistent with research on emotions, sex role stereotypes also often include the idea that females are more empathic than males (see Basow, 1986). Various measures of sex stereotypes reflect beliefs that females are more aware of others’ feelings, are more concerned about others’ welfare, and are more caring and nurturing (e.g., Bem, 1974; Broverman et al., 1972; Eagly, 1987; Eisenberg & Lennon, 1983; Spence & Helmreich, 1978).
Although sex differences in emotion and empathy generally conform to social stereotypes, it is clear that considerable variation exists within each sex. That is, while females as a group may score higher than males on some measures of emotional responsiveness, some females will likely score lower than males and some males will likely score higher than females. These variations imply that some variable other than biological sex may more accurately predict sex differences commonly observed in the literature.
Gender Role Self-Perception
Gender role self-perception is one particularly promising variable that may help in further explaining within-sex and between-sex variations in empathy and affective responses. Specifically, Eagly (1987; also see Bakan, 1966) suggests that many aspects of social behavior are likely the result of social expectations (i.e., attitudes and beliefs) regarding appropriate male and female behaviors (also see Bem, 1993; Epstein, 1988). Eagly explained that these gender-linked expectations can be conceptualized along two dimensions, each of which define positive
personal attributes. The communal dimension, typically associated with females, consists of the aforementioned qualities such as nurturance, empathy, and emotionality. The agentic dimension, often associated with males, consists of qualities such as independence and self-assertion. Eagly suggests that because gender role expectations are normative and shared, many individuals are likely to exhibit social perceptions, emotional reactions, and behaviors that are consistent with these expectations. Furthermore individuals internalize expectations and conform to gender role norms in the absence of any social or institutional pressures. Consequently, under some circumstances, an individual’s psychological gender, or the internalization of communal or agentic qualities, may be a better predictor of responses than an individual’s biological sex (see Bem, 1974, 1985; Frable, 1989; Spence, 1984).
Implications of Sex and Gender Role on Responses to Entertainment
Although numerous individual characteristics likely affect emotional responses, a large body of research suggests that empathic tendencies play important roles in viewers’ experience of entertainment (see Zillmann, 1991; Tamborini & Mettler, 1990). Individuals who evidence high levels of concern for others, experience the emotions of others in their environment, and who can easily imagine themselves from the perspectives of real or fictional characters should be particularly likely to experience intense emotional responses to dramatic portrayals. Consistent with these expectations, several researchers have reported that numerous dimensions of empathy such as empathic concern, fictional involvement, and perspective taking are predictive of emotional responses to horror films and to tear-jerkers (e.g., Davis, Hull, Young, & Warren, 1987; Oliver, 1993; Tamborini & Mettler, 1990; Tamborini, Stiff, & Heidel, 1990).
Given that the qualities typically associated with females include many of these empathic dimensions, as well as increased propensity for emotionality per se, it follows that females should be more likely than males to exhibit responses consistent with empathetic reactions. That is, females should be particularly likely to report identifying with characters’ problems and feeling concern for their well-being. In addition, prior research on responses to specific film genres suggests that males’ and females’ levels of distress and enjoyment should vary as a function of the type of film under consideration. This reasoning led to the following hypotheses:
H1: Females will report greater disturbance in response to distressing film scenes (violent or tragic) than will males.
H2: Females will report greater enjoyment of tragedy and less enjoyment of violence than will males.
H3: Females will report greater empathetic responsiveness to film characters than will males.
As noted earlier, however, sex may not be the most accurate reflection of internalization of gender-related norms. Consequently, it follows that an individuals’ psychological gender role serf-perception could be a better predictor of empathetic and emotional responding. This reasoning resulted in the following hypotheses:
H4: Communal participants will report greater disturbance in response to distressing film scenes (violent or tragic) than will agentic participants.
H5: Communal participants will report greater enjoyment of tragedy and less enjoyment of violence than will agentic participants.
H6: Communal participants will report greater empathetic responsiveness to film characters than will agentic participants.
These hypotheses led to the present investigation, which was conducted in two phases. In the first phase, male and female participants completed the Bern Sex Role Inventory (BSRI; Bem, 1976) as part of an in-class exercise. In the second phase, about six weeks later, the participants viewed three brief scenes from contemporary movies that involved either neutral, tragic, or violent content. Following exposure, the participants reported their affective reactions to each segment.
Female (n = 207) and male (n = 193) undergraduates participated in this study. These students were enrolled in an introductory-level professional communication course at a large university in the Southeastern United States. The course is a core requirement for the majority of undergraduate majors at the university and draws students from a variety of interests and disciplines. The ethnic breakdown of this population was approximately 93% Caucasian, 5% African American, and 2% other. All participants were volunteers and received course credit for their involvement in the study.
The first phase of the study was conducted as part of an in-class exercise. Participants completed the BSRI (Bem, 1976), which was incorporated with several other inventories into a self-administered questionnaire. The questionnaire was completed during the initial weeks of the academic quarter prior to any class discussion of gender role self-perceptions, empathic and emotional responding, or related concepts.
In the second phase of the research project, approximately six weeks later, those students who had completed phase one of the study were offered course extra credit to serve as research participants in a brief experiment exploring people’s responses to modern films. Those interested in the project were scheduled to attend one of several sessions held outside of class at various times over a one week period.
Participants were tested in same-sex groups of 10-13 persons by a female experimenter. The experimenter read a description of the project explaining that the participants would be viewing and evaluating three short segments from different feature films. Further, the participants were cautioned that some of the segments contained material of an adult nature and that anyone who might be offended by such materials was free to discontinue participation without penalty. None of the participants objected to viewing the materials.
Following these instructions, the participants were asked to read and complete an informed consent form which restated their right to withdraw from the experiment at any time. None withdrew at any time during the study. Subsequently, the experimenter distributed the film evaluation questionnaire, pointed out that it contained three identical sets of pages (i.e., one set for each film segment), and noted that the videotape would be stopped after each segment so that the participants could indicate their opinions. The experimenter then presented the stimulus materials. Following completion of the exposure treatment, the participants were thoroughly debriefed about the investigation and thanked for their participation.
Gender Role Assessment
The BSRI (Bem, 1976, 1985) includes 40 self-descriptive attributes. Participants were asked to indicate how well each attribute described herself or himself on a scale ranging from never or almost never true (0) to always or almost always true (4).
Incorporated in the inventory are 20 items that reflect our culture’s definition of femininity (e.g., tender, understanding) and 20 items that reflect masculinity (e.g., assertive, independent). In a manner adapted from previous research (cf. Frable & Bem, 1985) femininity (M = 50.20, SD = 8.18, a = .83) and masculinity (M = 54.66, SD = 8.66, a = .85) scales were computed by summing the responses for each participant. Examination of the correlation between the femininity and masculinity scales revealed a negligible (r = -0.10) but significant (p [less than] 0.05) relationship. Then, the femininity and masculinity scale scores were standardized within each sex, the femininity scale scores were reversed ([X.sup.*]-1), and the mean of the two scales was computed for each participant. Conceptually, this operationalization is the same as the “gender schematicity” typology advanced by Bern (1985) and her associates (Frable & Bem, 1985), but offers two advantages. First, both a continuo
us and discrete variable can be computed to operationalize gender role self-perceptions. This, of course, provides for greater versatility. Second, when forming the discrete variable, this approach allows for essentially equal cell sizes within each gender rather than the 25% sex-typed, 25% cross-sex-typed, and 50% undifferentiated division resulting from Bem’s scheme.
These transformations produced a single index of gender role self-perception that ranged from communal (negative values) to agentic (positive values) that effectively incorporated the information contained within the femininity (r = 0.71, p [less than] 0.0001) and masculinity (r = 0.71, p [less than] 0.0001) scales. A tertile distribution of this index was then computed in order to provide a discrete gender role type variable with three levels: communal (females, n = 65; males, n = 69), undifferentiated (females, n = 63; males, n = 70), and agentic (females, n = 65; males, n = 68). For subsequent tests, only responses from the communal and agentic participants (n = 267) were retained.
The exposure materials consisted of three segments, each approximately eight minutes long, taken from contemporary films. These films were chosen as prototypical examples of the genres of interest in this study: tragic films, violent films, and neutral films (or dramas). A segment from the film Beaches was used to operationalize “tragedy.” This segment depicted the relationship between two women, one suffering from terminal cancer, as they renew their childhood friendship. The cancer patient is shown in the final days of her life, making preparation for the care of her young daughter, and then dying. The very melodramatic segments ends with her funeral.
A “violent” film segment was operationalized using an excerpt from Friday, the 13th, Part III. The segment, which was typical of contemporary horror films (Weaver, 1991b; also see Sapolsky & Molitor, 1996), showed two young men and a woman chased and attacked by a murderous maniac, Jason. The violet segment ends with the three people defending themselves against Jason, but clearly on the verge of peril.
A “neutral” segment was taken from the movie Yentil. The segment portrayed a young woman walking through the marketplace of a 19th century European village. Interested in pursuing academic interests during a time when few women were involved in education, the young woman is shown experiencing difficulty purchasing a book. The scene ends, however, with the young woman receiving tutoring in reading by her father.
The film segments were presented in a fashion designed to minimize any order effect. Specifically, the neutral segment was viewed and evaluated first by the participants in all exposure groups. Then, the tragedy and violent segments were presented in a balanced order (i.e., neutral, violent, tragedy or neutral, tragedy, violent). These two presentation orders were viewed by approximately equal numbers of participants within each sex.
After viewing each scene, the participants were asked to report their affective reactions to both the film segment and to the characters portrayed in the segment. Specifically, on a 5-point scale ranging from strongly disagree (0) to strongly agree (4), participants indicated their responses to a 15-item “Affective Reactions to Film” inventory. This inventory included five items to assess reactions to the films (cf. Mundorf et al., 1989) and 10 items tapping empathetic responses to the film characters (cf. Richendoller & Weaver, 1994; Tamborini & Mettler, 1990).
Based on these responses, three dependent measures (disturbance, enjoyment, empathetic responsiveness) were created for each film segment. Guided by a principal components factor analysis, two indices assessing affective reactions to the film segments were created. Responses to “The film segment I just viewed was disturbing” and “The film segment I just saw was distressing” were averaged to produce a “disturbance” index (M = 1.98, SD = 0.66, [Alpha] = 0.85). Two items – “The film segment I just viewed was entertaining” and “I found the film segment enjoyable” – were averaged to produce an “enjoyment” index (M = 1.97, SD = 0.70, [Alpha] = 0.79). A fifth item included in the factor analysis – “The film segment I just saw was delightful” – proved confounded on both factors and was excluded from subsequent analyses.
The 10 remaining items of the “Affective Reactions to Film” inventory were, following principal components factor analysis, combined to form a single empathetic responsiveness (M = 2.09, SD = 0.53, [Alpha] = 0.91) index. This measure was created by averaging responses to “I tended to get emotionally involved with the character’s problems,” “I really got involved with the feelings and characters in this video segment,” “Learning about the character’s misfortune made me feel sad,” “While watching this video segment, I felt partly as though I were one of the characters,” “I found it difficult to see things from the character’s point of view” (reversed), “I didn’t feel very sorry for the characters when they were having problems” (reversed), “When I saw the character(s) treated unfairly, I didn’t feel very much pity for them” (reversed), “I could very easily put myself in the place of the leading character,” “I tried to understand the characters better by imagining how things looked from their perspective,” and “While watching the video segment, I imagined how I would feel if the event in the story were happening to me.”
For each participant, a single score was computed on each index (disturbance, enjoyment, empathetic responsiveness) for each film segment. These measures constituted the dependent measures used in subsequent analyses.
The three dependent measures were each subjected to a 2 x 2 x 3 analysis of variance model for mixed measures with participant sex (female, male) and gender role (agentic, communal) as independent-measure factors and film type (neutral, tragic, violent) as the repeated-measure factor. In these analyses, tests for within-subjects effects were adjusted using the Huynh-Feldt epsilon (Huynh & Feldt, 1970). Subsequent mean comparisons were computed using Fisher’s LSD test. Following recommendations of Vasey and Thayer (1987; also see O’Brien & Kaiser, 1985), multivariate tests were also computed for the within-subjects main effect and the interactions using the Hotelling-Lawley algorithm. The statistical decisions produced by these tests were identical to those of the univariate procedures in all cases. Consequently, the multivariate tests are not reported.
The within-subjects tests on the disturbance measure revealed a significant main effect for film type [F(2,526) = 100.60, p [less than] 0.0001, [[Eta].sup.2] = 0.28], a Participant Sex x Film Type interaction [F(2, 526) = 16.32, p [less than] 0.0001, [[Eta].sup.2] = 0.06], and a Gender Role x Film Type interaction IF(2, 526) = 5.85, p [less than] 0.004, [[Eta].sup.2] = 0.02]. The Participant Sex x Gender Role x Film Type interaction proved negligible (F [less than] 1), however. The significance levels of the F tests were adjusted based on a Huynh-Feldt epsilon of 0.9835.
Table I reports the means for the film type main effect and for the Participant Sex x Film Type interaction. Although Hypothesis 1 predicted that females would report greater disturbance in response to the tragic and violent films, these results showed that females’ disturbance ratings were significantly higher than were males’ disturbance ratings for each film type. However, inspection of the means within same-sex groups reveals a slightly different pattern for male and female participants. For male participants, highest ratings of disturbance were obtained for the tragic segment. In contrast, females rated the tragic and violent segments as equally disturbing, and more so than the neutral segment.
Table I. Disturbance, Enjoyment, and Empathetic Responsiveness as a Function of Participant Sex and Film Type(a) Film Type Participant Sex Neutral Tragic Violent Disturbance Male (n = 135) 1.25a 2.44c 1.46ab Female (n = 130) 1.63b 2.69d 2.55cd Combined (n = 265) 1.43A 2.56C 1.99B Enjoyment Male 1.50a 1.80b 2.47c Female 1.73ab 2.50c 1.61ab Combined 1.61A 2.14B 2.05B Empathetic responsiveness Male 1.67b 2.48d 1.42a Female 2.09c 3.20e 1.57ab Combined 1.87B 2.83C 1.49A a For the Participant Sex x Film Type interaction effect, means with different lowercase letters differ at p [less than] 0.05 by Fisher's LSD test. For the film type main effect (participant sex combined) means having different uppercase letters differ at p [less than] 0.05 by Fisher's LSD test.
Table II reports the means associated with the Gender Role x Film Type interaction. Although both gender role groups reported the highest levels of disturbance in response to the tragic film, mixed support was obtained for Hypothesis 4. As predicted, communal participants rated the violent film as more disturbing than did the agentic participants. However, contrary to expectations, agentic participants rated the tragic film as more disturbing than did the communal participants.
Examination of the between-subjects univariate tests revealed only a participant sex main effect [F(1, 263) = 53.58, p [less than] 0.0001, [[Eta].sup.2] = 0.17]. The gender role main effect and the Participant Sex x Gender Role interaction produced only negligible variation (F [less than] 1). Examination of the means associated with the sex main effect reveals that females (M = 2.29) reported significantly more disturbance than did males (M = 1.72).
Table II. Disturbance and Enjoyment as a Function of Gender Role Self-Perception and Film Type(a) Film Type Gender Role Neutral Tragic Violent Disturbance Communal (n = 133) 1.38a 2.49d 2.18c Agentic (n = 132) 1.49a 2.64e 1.83b Enjoyment Communal 1.72ab 2.32d 1.95bc Agentic 1.52a 1.98c 2.13cd a Means with different lowercase letters differ at p [less than] 0.05 by Fisher's LSD test.
The within-subject tests on the enjoyment measure revealed a significant main effect for film type [F(2, 526) = 18.63, p [less than] 0.0001, [[Eta].sup.2] = 0.07], a Participant Sex x Film Type interaction [F(2, 526) = 37.82, p [less than] 0.0001, [[Eta].sup.2] = 0.13], and a Gender Role x Film Type interaction [F(2, 526) = 4.30, p [less than] 0.02, [[Eta].sup.2] = 0.02]. The three-way interaction proved negligible (F [less than] 1), however. The significance levels of the F tests were adjusted based on a Huynh-Feldt [Epsilon] of 0.9020.
Table I reports the means associated with the main effect for film type and with the Participant Sex x Film Type interaction. Male participants reported significantly more enjoyment after viewing the violent film segment than either the neutral or tragic segments. In contrast, female participants’ enjoyment ratings were highest for the tragic segment and lowest for the violent segment. Consistent with Hypothesis 2, sex differences in enjoyment of specific film segments were also apparent: Male participants reported greater enjoyment of the violent segment and less enjoyment of the tragic segment than did females.
Table II presents the means associated with the Gender Role x Film Type interaction. Consistent with Hypothesis 4, agentic participants rated the violent segment as most enjoyable, whereas communal participants rated the tragic segment as most enjoyable.
Finally, between-subjects univariate tests revealed only a significant Participant Sex x Gender Role Interaction [F(1, 263) = 10.36, p [less than] 0.002, [[Eta].sup.2] = 0.04]. The participant sex (F [less than] 1) and gender role (F = 2.25) main effects produced variation below conventional levels of significance. Table III presents the interaction effect means which reveal communal males reported higher enjoyment ratings than agentic males, whereas enjoyment ratings for communal and agentic females did not differ.
Table III. Enjoyment and Empathetic Responsiveness as a Function of Participant Sex and Gender Role Self-Perception(a) Gender Role Participant Sex Communal Agentic Enjoyment Male 2.11c 1.74a Female 1.88ab 2.01bc Empathic responsiveness Male 2.03b 1.68a Female 2.28c 2.29c a Means with different lowercase letters differ at p [less than] 0.05 by Fisher's LSD test.
The within-subjects tests on the empathetic responsiveness measure revealed a significant main effect for film type [F(2, 526) = 276.06, p [less than] 0.0001, [[Eta].sup.2] = 0.51] and a significant Participant Sex x Film Type interaction [F(2, 526) = 11.85, p [less than] 0.0001, [[Eta].sup.2] = 0.04]. The other interactions yielded only negligible variation (F [less than] 1). The significance levels of the F tests were adjusted based on a Huynh-Feldt [Epsilon] of 0.8651.
Table I reports the means associated with the main effect for film type and with the Participant Sex x Film Type interaction. Although both males and females reported the highest levels of empathetic responsiveness in reaction to the tragic film, females’ ratings were significantly higher than males on both the neutral and tragic segments, and generally so for the violent segment.
The between-subjects univariate tests revealed significant main effects for participant sex [F(1,263) = 52.56, p [less than] 0.0001, [[Eta].sup.2] = 0.17] and for gender role [F(1, 263) = 8.65, p [less than] 0.004, [[Eta].sup.2] = 0.03], and a significant Participant Sex x Gender-Role interaction [F(1, 263) = 9.11, p [less than] 0.003, [[Eta].sup.2] = 0.03]. Consistent with Hypotheses 3 and 5, females (M = 2.28) reported significantly more empathetic responsiveness than their male (M = 1.86) counterparts, and communals (M = 2.16) reported greater empathetic responsiveness than did agentics (M = 1.98). However, the main effect for gender role should be interpreted in light of the Participant Sex x Gender Role interaction. Examination of the means reported in Table III shows that females reported significantly higher empathetic responsiveness than did males regardless of gender role group. However, within same-sex groups, communal males reported higher levels of empathetic responsiveness than did agentic males, whereas empathetic responsiveness did not differ for communal and agentic females.
The results of this study highlight the importance of the viewer’s sex in emotional and empathetic responding to films, and further suggest that the viewer’s experience of media entertainment likely reflects a combination of the viewer’s sex, gender role self-perception, and the type of entertainment in question. Consistent with previous research, this study revealed numerous sex differences in responses to each type of entertainment explored. In terms of the tragic film, females reported greater disturbance, enjoyment, and empathetic responsiveness. These results replicate previous research reporting that females experience greater sadness in response to tear-jerkers than do males, but also simultaneously report greater enjoyment of this type of entertainment (Oliver, 1993). Also consistent with prior research, females in this study reported experiencing greater disturbance, less enjoyment, and tended to express greater empathetic responsiveness to the violent film than did males (see Cantor & Reilly, 1982; Sparks, 1986; Tamborini & Stiff, 1987; Zillmann et a
l., 1986). Finally, sex differences were obtained for the neutral film clip, with females reporting greater disturbance and empathetic responsiveness than males.
Many significant differences were also revealed between the communal and agentic gender role groups. Consistent with prior research, communal participants reported greater disturbance to the violent film and greater enjoyment of the tragic film. In addition, communal participants also reported higher levels of empathetic responsiveness overall than did agentic participants. However, this main effect for gender role must be interpreted in light of the interaction obtained between gender role and participant sex. The pattern revealed in this interaction suggests that gender role self perception has a stronger influence on males’ responses than on females’ responses to films. Among the male participants, communal participants reported greater levels of enjoyment and empathetic responsiveness; among the female participants, responses of communals and agentics did not differ.
Finally, contrary to expectations, communal participants reported less disturbance to the tragic film than did the agentic participants. These results are particularly surprising given that communal orientations reflect a tendency toward warmth, caring, and affiliation (Eagly, 1987; Spence & Helmreich, 1978). However, it is important to note that agentic participants rated the tragic film as less enjoyable than did the communal participants. Although this pattern of results may suggest that enjoyment of sad films is negatively associated with disturbance, prior research has reported positive correlations between ratings of enjoyment and ratings of sadness for specific films in this genre (Oliver, 1993). Clearly, future research in this area would benefit from exploring the complex relationships between viewers’ reactions that include both positive and negative emotional responses.
Although this study obtained support for many of the expected sex and gender role differences in responses to media entertainment, several additional unexpected findings deserve further attention. First, whereas the reported sex differences in response to the tragic and violent films were consistent with expectations, it is curious why sex differences were also obtained for the neutral film where no clear gender-linked emotional responses were portrayed. Perhaps findings for the neutral film reflect a greater general tendency for emotional responsiveness among females than among males (see Brody & Hall, 1993, for a review). Although this interpretation is consistent with prior research on sex differences in the experience and display of emotions, it would also suggest the counterintuitive prediction that females should respond more strongly to genres such as comedy, science fiction, or action adventure.
An alternative explanation for these sex differences may involve some effect of the type of characters portrayed in the neutral film. Although the film clip employed in this study included both male and female characters, the gender-polarized experience of one female character was particularly salient. Viewing with a contemporary eye, females may have more easily identified with the protagonist’s blight and, therefore, may have responded more intensely to her emotions (Feshbach & Roe, 1968; Hoffner & Cantor, 1991; Tannenbaum & Gaer, 1965). This interpretation suggests that future research may benefit from exploring the role of both the viewers’ sex and the sex of the main characters in predicting emotional responses to media entertainment (see Oliver, Weaver, & Sargent, 1997). Indeed, this interpretation may help explain the consistent and robust sex differences in responses to tear-jerkers that typically explore the plight of female characters (e.g., Terms of Endearment, Steel Magnolias, Fried Green Tomatoes, etc.).
Another finding that deserves additional attention concerns the particularly strong influence of gender role self perception for male participants only. Indeed, these results imply that gender roles have essentially no influence on females, but have a very strong influence on males. Although this broad interpretation seems at odds with prior research on behavioral and attitudinal correlates of gender identity (see Cook, 1985), these findings may be partially explained in terms of the avoidance of “inappropriate” behaviors. Specifically, this study examined the expression of emotional and empathetic reactions. Given that emotionality and empathetic responsiveness are stereotyped as more appropriate for females than for males, these findings may reflect sex-typed males’ (agentics) avoidance of “feminine” behaviors. According to Bem (1985),” . . .the sex-typed individual’s pattern of avoiding gender-inappropriate activity is not restricted to relatively trivial activities but extends to complex social behaviors like independence and nurturance” (p. 210). It should be noted, however, that this explanation fails to account for the lack of gender role effects for males’ ratings of disturbance. In addition, if this “avoidance” interpretation fully accounted for participants’ reactions, one might expect to see more pronounced interactions between sex, gender roles, and film types. For example, this explanation would likely predict that communal females should be particularly unlikely to report enjoyment of the violent film. Certainly, the complexity of asymmetrical gender role correlates has been noted elsewhere (see Bem, 1987) and is worthy of further consideration.
The influences of sex and gender role serf-perceptions obtained in this study emphasize the importance of these variables on viewers’ emotional responses to media entertainment. Although these variables are only a few of the possibly infinite factors that affect why one person may react with great emotional intensity to a film while the next person remains unmoved, these variables clearly deserve further attention. Perhaps by more closely examining the correlates of gender role self-perceptions and the media portrayals that compliment gender role expectations, researchers will be in a better position to disentangle the reasons why men and women often differ in their selection of and responses to media entertainment.
Allen, J. G., & Haccoun, D. M. (1976). Sex differences in emotionality: A multidimensional approach. Human Relations, 29, 711-722.
Antill, J. R. (1987). Parents beliefs and values about sex roles, sex differences, and sexuality: Their sources and implications. Review of Personality and Social Psychology, 7, 294-328.
Bakan, D. (1966). The duality of human existence: An essay on psychology and religion. Chicago: Rand McNally.
Basow, S. A. (1986). Gender stereotypes: Traditions and alternatives (2nd ed.). Pacific Grove, CA: Brooks/Cole.
Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 155-162.
Bem, S. L. (1976). Probing the promise of androgyny. In A. G. Kaplan & J.P. Bean (Eds.), Beyond sex-role stereotypes: Readings toward a psychology of androgyny. Boston: Little Brown.
Bem, S. L. (1985). Androgyny and gender schema theory.’ A conceptual and empirical integration. In T. B. Sonderegger (Ed.), Nebraska symposium on motivation: Vol. 32, Psychology of gender. Lincoln: University of Nebraska Press.
Bem, S. L. (1987). Probing the promise of androgyny. In: M. R. Walsh (Ed.), The psychology of women: Ongoing debates. New Haven, CT:. Yale University Press.
Bem, S. L. (1993). The lenses of gender. New Haven, CT: Yale University Press.
Birnbaum, D. W., Nosanchuk, T. A., & Croll, W. L. (1980). Children’s stereotypes about sex differences in emotionality. Sex Roles, 6, 435-443.
Brody, L. R. (1993). On understanding gender differences in the expression of emotion: Gender roles, socialization, and language. In S. L. Ablon, D. Brown, E. J. Khantzian, & J. E. Mack (Eds.), Human feelings: Explorations in affect development and meaning. Hillsdale, NJ: The Analytic Press.
Brody, L. R., & Hall, J. A. (1993). Gender and emotion. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions. New York: Guilford.
Broverman, I., Vogel, S. R., Broverman, D. M., Clarkson, F. E., & Rosenkrantz, P. S. (1972). Sex role stereotypes: A current appraisal. Journal of Social Issues, 28, 59-78.
Cantor, J., & Reilly, S. (1982). Adolescents’ fright reactions to television and films. Journal of Communication, 32, 87-99.
Cook, E. P. (1985). Psychological androgyny. New York: Pergamon Press.
Davis, D. M. (1990). Portrayals of women in prime-time network television: Some demographic characteristics. Sex Roles, 23, 325-333.
Davis, M. H., Hull, J. G., Young, R. D, & Warren, G. G. (1987). Emotional reactions to dramatic film stimuli: The influence of cognitive and emotional empathy. Journal of Personality Psychology and Social Psychology, 52, 126-133.
Eagly, A. H. (1987). Sex differences in social behavior: A social-role interpretation. Hillsdale, NJ: Lawrence Erlbaum Associates.
Eisenberg, N., & Lennon, R. (1983). Sex differences in empathy and related capacities. Psychological Bulletin, 94, 100-131.
Epstein, C. F. (1988). Deceptive distinctions: Sex, gender, and the social order. New Haven, CT: Yale University Press.
Feshbach, N. D., & Roe, K. (1968). Empathy in six- and seven-year-olds. Child Development, 39, 133-145.
Frueh, T., & McGhee, P. E. (1975). Traditional sex role development and amount of time spent watching television. Developmental Psychology, 111, 109.
Frable, D. E. S. (1989). Sex typing and gender ideology: Facets of the individual’s gender psychology that go together. Journal of Personality and Social Psychology, 56, 95-108.
Frable, D. E. S., & Bem, S. L. (1985). If you are gender schematic, all members of the opposite sex look alike. Journal of Personality and Social Psychology, 49, 459-468.
Frieze, I. H., Parsons, J. E., Johnson, R B., Ruble, D. N., & Zellman, G. L. (1978). Women and sex roles: A social psychological perspective. New York: Norton.
Gerbner, G., Gross, L, Morgan, M., & Signorielli, N. (1986). Living with television: The dynamics of the cultivation process. In J. Bryant & D. Zillmann (Eds.), Perspectives on media effects. Hillsdale, NJ: Lawrence Erlbaum.
Hoffner, C., & Cantor, J. (1991). Perceiving and responding to mass media characters. In J. Bryant & D. Zillmann (Eds.), Responding to the screen: Reception and reaction processes. Hillsdale, NJ: Lawrence Erlbaum Associates.
Huynh, H., & Feldt, L. S. (1970). Conditions under which mean square ratios in repeated measurement designs have exact F-distributions. Journal of the American Statistical Association, 65, 1582-1589.
Jeffres, L. W. (1994). Mass media processes (2nd ed.). Prospect Heights, IL: Waveland.
Lombardo, W. K., Cretser, G. A., Lombardo, B., & Mathis, S. (1983). Fer cryin’ out loud – There is a sex difference. Sex Roles, 9, 987-995.
Mundorf, N., Weaver, J., & Zillmann, D. (1989). Effects of gender roles and self perceptions on affective reactions to horror films. Sex Roles, 20, 655-673.
O’Brien, R. G., & Kaiser, M. K. (1985). MANOVA method for analyzing repeated measures designs: An extensive primer. Psychological Bulletin, 97, 316-333.
Oliver, M. B. (1993). Exploring the paradox of the enjoyment of sad films. Human Communication Research, 19, 315-342.
Oliver, M. B., Weaver, J. B., III, & Sargent, S. L. (1997). An Examination of Factors Related to Sex Differences in Enjoyment of Sad Films. Paper submitted for publication, Auburn University.
Richards, J., & Sheridan, D. (1987). Mass observation at the movies. London: Routledge & Kegan.
Richendoller, N. R., & Weaver, J. B., III (1994). Exploring the links between personality and empathic response style. Personality and Individual Differences, 17, 303-311.
Sapolsky, B. S., & Molitor, F. (1996). Content trends in contemporary horror films. In J. B. Weaver, III & R. Tamborini (Eds.), Horror films: Current research on audience preferences and reactions. Hillsdale, NJ: Lawrence Erlbaum Associates.
Shields, S. A. (1987). Women, men, and the dilemma of emotion. In P. Shaver & C. Hendrick (Eds.), Review of personality and social psychology: Vol. 7. Sex and gender. Newbury Park, CA: Sage.
Sparks, G. G. (1986). Developing a scale to assess cognitive responses to frightening films. Journal of Broadcasting and Electronic Media, 30, 65-73.
Spence, J. (1984). Masculinity, femininity, and gender-related traits: A conceptual analysis and critique of current research. In B. A. Maher & W. B. Maher (Eds.), Progress in experimental personality research (Vol. 13). New York: Academic Press.
Spence, J. T., & Helmreich, R. L. (1978). Masculinity & femininity: Their psychological dimensions, correlates, & antecedents. Austin: University of Texas Press.
Staley, A. A., & O’Donnell, J.P. (1984). A developmental analysis of mothers’ reports of normal children’s fears. Journal of Genetic Psychology, 144, 165-178.
Tamborini, R., & Mettler, J. (1990, November). Emotional reactions to film: A model of empathic processes. Paper presented at the annual conference of the Speech Communication Association, Chicago, IL.
Tamborini, R., & Stiff, J. (1987). Predictors of horror film attendance and appeal: An analysis of the audience for frightening films. Communication Research, 14, 415-436.
Tamborini, R., Stiff, J., & Heidel, C. (1990). Reacting to graphic horror: A model of empathy and emotional behavior. Communication Research, 17, 616-640.
Tamborini, R., Stiff, J., & Zillmann, D. (1987). Preference for graphic horror featuring male versus female victimization: Personality and past film viewing experiences. Human Communication Research, 13, 529-552.
Tannenbaum, P. H., & Gaer, E. P. (1965). Mood change as a function of stress of protagonist and degree of identification in a film-viewing situation. Journal of Personality and Social Psychology, 2, 612-616.
Thompson, T. L, & Zerbinos, E. (1995). Gender roles in animated cartoons: Has the picture changed in 20 years? Sex Roles, 32, 651-673.
Vasey, M. W., & Thayer, J. F. (1987). The continuing problem of false positives in repeated measures ANOVA in psychophysiology: A multivariate solution. Psychophysiology, 24, 479-486.
Weaver, J. B., III (1991a). Exploring the links between personality and media preferences. Personality and Individual Differences, 12, 1293-1299.
Weaver, J. B., III (1991b). Are “slasher” horror films sexually violent? A content analysis. Journal of Broadcasting and Electronic Media, 35, 385-392.
Weaver, J. B., III, Brosius, H. B., & Mundorf, N. (1993). Personality and movie preferences: A comparison of American and German audiences. Personality and Individual Differences, 14, 307-315.
Zillmann, D. (1991). Empathy: Affect from bearing witness to the emotions of others. In J. Bryant & D. Zillmann (Eds.), Responding to the screen.’ Reception and reaction processes. Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D., & Weaver, J. B., III (1996). Horror, gender roles, and the socialization of affect. In J. B. Weaver, Ill & R. Tamborini (Eds.), Horror films: Current research on audience preferences and reactions. Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D., Weaver, J. B, Mundorf, N., & Aust, C. F. (1986). Effects of an opposite-gender companion’s affect to horror on distress, delight, and attraction. Journal of Personality and Social Psychology, 51, 586-594.
COPYRIGHT 1998 Plenum Publishing Corporation
COPYRIGHT 2000 Gale Group