The Shining or The Sixth Sense

Young Men’s and Women’s Different Autobiographical Memories of the

Experience of Seeing Frightening Movies on a Date

Richard Jackson Harris

Steven J. Hoekstra

Christina L. Scott

Fred W. Sanborn

Joseph Andrew Karafa

Jason Dean Brandenburg Kansas State University

The present study used autobiographical memory to investigate the social experience and short- and long-term effects of seeing frightening movies on a date, extending Zillmann and Weaver’s (1996) model of differential gender- role behaviors to persons’ own real-life dating experiences. Young adult participants (a) recalled the experience of watching a scary movie on a date, and (b) were assessed for levels of gender-role traditionality, sensation seeking, and dispositional empathy. Results showed that almost all individuals could recall such a date. Although men reported more positive reactions to the film and women more negative reactions, the experience appeared to have some social utility for both. Sex was a better predictor than the gender-role measures for Negative Reactions, Sleep Disturbances, and the likelihood of being Scared Today by the movie. Sensation-Seeking and Empathy were modest predictors of the same variables. In sum, the dating context seemed to encourage both men and women to behave and react in highly gender- stereotypical ways.

Requests for reprints should be sent to Richard J. Harris, Department of Psychology, Bluemont Hall 492, Kansas State University, 1100 Mid-Campus Drive, Manhattan, KS 66506–5302. E-mail: [email protected]

MEDIAPSYCHOLOGY, 2, 245–268. Copyright © 2000, Lawrence Erlbaum Associates, Inc.

Sitting together at the edge of their seats in a dark theater, a teenage couple watches a horror film. With hearts racing, palms sweating, and an occasional gasp of shock and terror, the couple is transfixed by scenes of people being shot, maimed, and decapitated. This seemingly gruesome and violent event is actually a quite common dating scenario for young adults in Western society. What remains to be seen is how this experience is encoded and recalled by young viewers and what lasting effects and social motivations underlie the viewing of these violent films.

Considerable research has documented negative effects of viewing filmed violence, including modeling, desensitization, and cultivation (see Donnerstein & Smith, 1997; Dubow & Miller, 1996; Gunter, 1994; Harris, 1999; Strasburger, 1995, for reviews). One of the clearest negative effects of viewing violence and horror is fear (Cantor, 1991, 1994, 1996, 1998a, 1998b; Hoff n e r, 1997; Tamborini, 1996). People are more fearful after watching violence, although affective and behavioral reactions differ by age and sex. Also, there are several other potential mediating variables that influence the emotional reactions to violent films, including the degree of empathy (Tamborini, 1996; Zillmann, 1991), affective disposition (Zillmann, 1998), and traditionality of gender-role identification (Mundorf, We a v e r, & Zillmann, 1989; Zillmann, We a v e r, Mundorf, & Aust, 1986).

Although considerable research on the effects of media violence has been conducted, less attention has been paid to the social environment of watching and viewers’ personality characteristics. The present study explored individuals’ memories for an experience of watching a frightening movie while on a date. The dating situation was used for two reasons. First, dating is one of the most common contexts in which teens and young adults view movies, and horror films are an immensely popular film genre among teens and young adults. Second, the dating context should allow for the strongest test of possible gender-role differences, predicted by Zillmann and Weaver’s (1996) model. Given that the heterosexual dating context would be expected to increase the salience of gender differences, one’s identity as a man or as a woman should be central, making dating an appropriate scenario for examining sex differences.

Additionally, building on findings of past research, correlations between various measures of such memories and the individual difference measures of dispositional empathy, sensation seeking, and gender-role traditionality were examined. Finally, we explored the degree to which memories for being frightened predicted current preference for watching similar movies and the belief that one would be scared by such a movie today.

2 4 6 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

MODELS OF ENJOYMENT OF VIOLENCE

The question of why people view horror and graphic violence has been examined in both the psychological and communication literatures. Individual differences in gender, dispositional level of violence, risk taking, disinhibition, sensation seeking, curiosity about what is taboo or “forbidden,” susceptibility to boredom, and the ability to emotionally distance oneself have all been associated with a liking of horror films (Goldstein, 1998; Lawrence & Palmgreen, 1996). According to the uses and gratifications perspective of media effects (A. M. Rubin, 1994), the particular reasons one watches graphic horror and the benefits one receives will, in part, determine (a) whether the viewer identifies with the aggressor or the victim and (b) the effects that are experienced (Johnston, 1995). For example, those who watch movies with a predominantly violent content because they like to see the “gore” are less likely than others to experience high levels of fear or state empathy (Johnston, 1995). In addition, social cues such as the identification with the coviewing group or the use of violent viewing for purposes of social control, mood management, or emotional expression have all been found to affect enjoyment of violent media (Goldstein, 1998).

The viewing of horror films in the context of a date highlights the social function of gender roles. Zillmann and Weaver (1996) proposed a model whereby reactions to viewing horror (and the display of those reactions) serve critical social functions in the adolescent dating process. More specifically, they suggested that watching horror films can act as a kind of rite of passage for adolescent males and females, in which gender-acceptable behaviors can be displayed. Although Zillmann and Weaver’s model contains several propositions, two are particularly relevant here. For example, one proposition states that boys learn to display fearlessness and protective competence, whereas girls learn to display fearfulness and protective need. Another proposition states that successfully performing these gender-specific displays is experienced as pleasurable, whereas failing to experience them is unpleasant. The model predicts that each sex reinforces these behaviors in the other, and anyone whose behavior violates these norms is less well liked by both sexes.

In an experimental test of the model, males enjoyed viewing the horror film Friday the 13th Part III much more in the company of a fearful female confederate than in the company of a fearless one, whereas females enjoyed the movie more with a fearless male than a fearful one (Zillmann, Weaver, Mundorf, & Aust, 1986). One purpose of the present study was to extend these findings to the more ecologically valid setting of the participants’own dating experiences.

MEMORIES FOR SCARY M O V I E S 2 4 7

Zillmann and Weaver (1996) also suggested that future research should examine trait variables that might predict enjoyment of horror films. Two such traits, empathy and sensation seeking, were included in the present research. Lazarus (1991) and Tamborini (1996) noted that empathy is a frequent negative correlate of horror film enjoyment. Empathy is conceptualized most often as a multidimensional construct involving the degree to which an individual can relate to, or vicariously feel, the emotional experience of another person. The present study conceptualized and measured empathy using the model of Davis (1983; Davis, Hull, Young, & Warren, 1987). This model has four components: Perspective Taking, Fantasy Empathy, Empathic Concern, and Personal Distress. Perspective Taking is the ability to see things from others’ viewpoints or to imagine what it would be like in their situation (Tamborini, 1996). Fantasy Empathy is the ease of becoming emotionally involved with fictional characters. Empathic Concern is the sensitivity and compassion to others’ misfortunes in response to seeing their distress, and Personal Distress is one’s own discomfort or aversion in response to another’s distress (Davis, 1983). These factors were chosen as they seemed potentially relevant for the media experience, in that all can be involved in the viewing of media violence.

The trait measure of sensation seeking may also be viewed as a multidimensional construct, following the conceptualization of Zuckerman (1994). Like the empathy construct, four components are measured independently: Thrill and Adventure Seeking, Experience Seeking, Disinhibition, and Boredom Susceptibility. This model looks not only at thrill seeking directly but also the dimension of experience seeking more generally, as well as one’s proneness to boredom and disinhibition. Any of these dimensions might predict enjoyment of watching media violence.

AUTOBIOGRAPHICAL MEMORY

If one is to investigate Zillmann and Weaver’s (1996) model in an ecologically valid way, a logical place to start is in a situation in which participants view real movies on real dates. However, practical and ethical concerns make direct observation problematic. Therefore, autobiographical memory, the recollection of events in a person’s own life, was used in the present study. In spite of the proliferation of autobiographical memory research in the last decade (Brewer, 1995; Conway, Bruce, & Sehulster, 1998; Ross, 1991; D. C. Rubin, 1986, 1995), until very recently there had been few, if any, studies of memories for media experiences. However, two recent studies (Harrison & Cantor, 1999; Hoekstra, Harris, & Helmick, 1999) examined autobiographical memory for

2 4 8 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

horror/suspense movies and demonstrated the usefulness of this methodology. Although the studies were undertaken independently, both asked young adults to recall an incident from their childhood or teen years during which they had viewed a movie that had seriously frightened and disturbed them “so much that the emotional effect endured after the program or movie was over” (Harrison & Cantor, 1999, p. 102). Both studies used a variety of open-ended free recall and quantitative scaling measures to assess the respondents’reactions to the films and their memory for the effects over time. Perhaps the most striking result was that in both studies almost all respondents had such a memory and could describe it vividly. Some frequently reported effects included sleep disturbances, specific and nonspecific fears, and mental preoccupation with stimuli from the film. Many participants reported the effects to be long lasting, in many cases still existing years afterward. In the Hoekstra et al. (1999) study, the median age at which the frightening movie had been seen was 11 years old. This was an average of 8 years prior to the time at which the experiment was conducted. These studies demonstrated the usefulness of autobiographical memory as a methodology for examining long-term effects of consuming violent media.

Autobiographical memories are, by nature, highly personal and affective, attributes that are also highly characteristic of memories of viewing frightening films (Conway et al., 1998). Although the retrospective nature of autobiographical memory naturally raises some validity issues, this approach nonetheless offers a way to examine very long-term effects without the problems associated with a longitudinal study. Although there have been hundreds, perhaps thousands, of studies on the effects of media violence, only a handful have addressed effects lasting longer than a few weeks or months. Autobiographical memory is a useful complement to other experimental and field methodologies examining shorter term effects of media violence, because it is able to assess memories often going back as far as several years. Also, much of the past research has been done on children and teens, but there has been limited research on the enduring effects that viewing violence at a young age may have on adult experience. In an attempt to move in that direction, the present research gathered data about the nature of adults’ memories of watching frightening movies seen on a date as a teen or young adult. In addition, the fact that participants had already seen these violent films on their own avoids the ethical question of showing underage participants disturbing materials that have been previously shown to have strong antisocial effects.

MEMORIES FOR SCARY M O V I E S 2 4 9

2 5 0 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

METHOD

Participants

The participants were 233 introductory psychology students (125 males, 108 females, 91% White, mean age = 19.2 years) from a large U.S. public university with open admissions. Students earned partial fulfillment of a course requirement for participating and were tested in groups of 15–30 in Fall 1998.

Materials and Procedure

Initially, participants were told that they would be completing an anonymous questionnaire about movies and their memories about the films. They were told to take as much time as they needed to think about and to complete the packet of materials of three personality scales and the questions about the movie experience.

Participants first completed the 24-item Personal Attributes Questionnaire (PAQ; Spence & Helmreich, 1978), which assessed specific behaviors related to sex roles. This measure was used to test the strength of gender roles in predicting viewer memories. The PAQ consists of 5-point scales on which participants describe themselves on various dimensions (e.g., not at all emotional–very emotional, very rough–very gentle, never cries–cries very easily). Its M scale consists of traits associated with stereotypical desirability for men, and the F scale consists of traits associated with stereotypical desirability for women. The MF (androgyny) scale consists of traits desirable for both men and women.

Next, participants completed a sensation-seeking questionnaire (Form V, Zuckerman, 1994) that consisted of 40 pairs of forced-choice items in which participants checked “the choice that more closely describes your likes or the way you feel.” It consists of four subscales: Thrill and Adventure Seeking (e.g., “I sometimes like to do things that are a little frightening”), Experience Seeking (e.g., “I like to try new foods that I have never tasted before”), Disinhibition (e.g., “I like wild, uninhibited parties”), and Boredom Susceptibility (e.g., “I can’t stand watching a movie that I’ve seen before”).

The last personality inventory was the 28-item Interpersonal Reactivity Index (IRI) (Davis, 1983; Davis et al.,1987). This is a composite measure of empathy, with seven items for each of four subscales: Empathic Concern (e.g., “I am often quite touched by things I see happen”), Perspective Taking (e.g., “Before criticizing somebody, I try to imagine how I would feel if I were in their place”), Fantasy Empathy (e.g., “I really get involved with the feelings of a character in

a novel”), and Personal Distress (e.g., “Being in a tense emotional situation scares me”). Participants were asked to indicate on a 5-point scale the degree to which each sentence described them. To avoid acquiescence bias, some items were written negatively and reverse-scored. Scores for each empathy component were computed by adding scores on items within each subscale.

Following the IRI, in order to ensure that the nature of the movie did not affect the personality ratings, each participant was asked to think about “a date that you went on as a teen or young adult where you watched a movie that frightened you.” Participants were then asked several questions about the content of the movie and their emotional reactions to the experience. They were allowed as much time as necessary, although most required only 20 to 30 minutes to complete the questions. Questions took a variety of formats, from rating scales to fill-in-the-blanks to brief descriptions of content. The specific format of particular questions is described in more detail in the presentation of the results.

RESULTS

Plan of Analyses

We first present a descriptive analysis of the data, looking at differences in how males and females reported and experienced the frightening movie on a date. Next, several individual-difference variables are examined, to test their ability to predict certain dependent variables.

Descriptive Data

Watching a frightening movie on a date was almost a universal experience in the sample. Only three participants (1.3%) were unable or unwilling to report a memory of viewing a scary movie on a date; these data were discarded.1 This finding complements those of Harrison and Cantor (1999) and Hoekstra et al. (1999), who found that almost all college students had memories of being scared by a movie as a child or teen, but now it also seems clear that at least one common social context for this experience was on a date.

In terms of the films seen, by far the most common were Scream and its sequel Scream 2, which together accounted for 37% of the responses. A distant second was I Know What You Did Last Summer, which, along with its sequel I Still Know What You Did Last Summer, comprised 6.4% of the responses. Next most frequent were Candyman, Event Horizon, and the Halloween series (3%–4% each). All other movies were reported by six or fewer respondents, and most by only one or two. With very few exceptions, the movies were all R-rated.

MEMORIES FOR SCARY M O V I E S 2 5 1

The mean age of the participants was 19.4 years for males and 19.5 years for females. Participants’estimated mean age at the time of viewing the movie was 16.7 years for males and 17.6 years for females. The large majority of films were seen between the ages of 16 and 19 (70% of males and 82% of the females).

Circumstances of Viewing

Respondents were given four choices to indicate the time of day, and most reported watching the film either in the evening (38%) or late at night (58%). Only 3% watched in the afternoon and less than 1% in the morning, with no sex differences. About half the time the films were seen in a theater and the other half were viewed in the participant’s home, the participant’s date’s home, or the home of a third party (see Table 1).

TABLE 1 Viewing Context (%)

Coviewers % Whose Choice? % Where? %

Males Date Only 48 You 18 Your Home 30 Another Couple 9 Your Date 15 Date’s Home 10 Group 34 Another Man 5 Another Home 15 Family 6 Another Woman 1 Theater 42

Group Choice 47 Don’t Remem. 13

Females Date Only 36 You 13 Your Home 19 Another Couple 12 Your Date 11 Date’s Home 9 Group 44 Another Man 2 Another Home 18 Family 6 Another Woman 1 Theater 53

Group Choice 63 Don’t Remem. 10

Note. Data based on multiple-choice responses. Percentages may not total 100 due to missing responses and rounding error.

On a question about who had chosen the movie, the largest numbers of both males and females checked the option “joint or group decision,” and the most

2 5 2 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

frequent coviewers were either “a group of people” (39%) or “no one else” [except date] (42%); see Table 1. When the movie was selected by an individual, there were no significant sex differences, indicating that males and females were equally responsible for choosing the film. Whereas 88% of the males and 66% of the females checked “I watched the movie from start to finish with full attention,” 31% of the females but only 7% of the males checked “I looked away a lot or left the room and returned a lot.” Only 4% of the females and 2% of the males reported “leaving without watching the entire movie.”

Participants’ reports of their behavior and reactions while they watched the movie and their reactions after the movie appear in Tables 2 and 3, partitioned by sex. For these questions, participants were presented with a list of these effects or concurrent behaviors and reactions and asked to circle all that they experienced during or after viewing the film. In terms of effects experienced after viewing (Table 2), over half of the sample remembered being generally anxious and fearful, whereas many others reported a wild imagination, specific fears related to the movie, and various sleep problems. On five of these seven anxiety and sleep problems, females reported them significantly more frequently than did males.

TABLE 2 Percent Reporting Effects after Viewing

Effect Experienced Males Females t(228)

General Fear/Anxiety 44 61 2.45* Wild Imagination 1 34 2.21* Fear of Sleeping Alone 8 19 2.52* Specific Fear 7 18 2.57* Sleeping w/ Lights On 2 11 3.05** Insomnia 6 4 ns Serious Nightmares 3 6 ns

Note. Data based on items checked in response to “Which effects do you remember experiencing after seeing the movie, either immediately, or sometime later?” *p < .05. **p < .01.

For the concurrent reactions and behaviors (Table 3), females reported themselves “generally very jumpy” and “disgusted” more than males did, and they more often checked that they “held on to my date,” “hid my eyes or looked

MEMORIES FOR SCARY M O V I E S 2 5 3

away,” and “yelled and screamed.” On the other hand, males more often reported being “amused and entertained,” “sexually turned on,” and “surprised at my date’s reactions.” Because of the possibility of experimenter-wise alpha error from the large number of independent t tests, a level of .01 was used for this set of analyses.

TABLE 3 Percent Reporting Behaviors/Reactions Concurrent with Viewing

Behaviors/Reactions Experienced Males Females t(228)

Heart Beat Fast 53 67 ns Felt Amused and Entertained 59 41 3.06** Was Generally Very Jumpy 31 56 3.85*** Laughed 45 32 ns Held Onto My Date 21 55 5.54*** Hid Eyes/Looked Away 5 46 8.29*** Tried Not to Show How Scared I Was 24 25 ns Yelled and Screamed 6 32 5.13*** Felt Disgusted 10 26 2.60** Stomach Felt Funny 12 17 ns Surprised at Date’s Reactions 22 6 3.41*** Felt Sexually Turned On 21 4 4.08*** Thought the Movie Was Stupid 4 8 ns Was Shaking 3 6 ns Upset at Date’s Reactions 6 1 ns Cried 2 1 ns

*p < .05, **p < .01, ***p < .001

Based on an exploratory factor analysis and the findings of previous research, the combination of conceptually similar resulting factors 2 clustered the concurrent behaviors and reactions into three composite variables: Positive Reactions, Negative Reactions, and Sleep Disturbances, each formed by adding the number of the component items circled by the participant. The mean totals for each of these three variables appear in Table 4. Overall, males reported almost twice as many Positive Reactions as females, whereas females reported about twice as many Negative Reactions and Sleep Disturbances (all ps < .001). Although these differences were quite robust, some caution is in order because

2 5 4 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

of low reliabilities of the three composite scales (Cronbach’s α between .40 and .56). Thus the more important results may be considered to be the individual behavior differences in Tables 2 and 3. All in all, however, results suggest that males and females together on a date are experiencing these movies very differently.

TABLE 4 Mean Number of Responses on Three Composite Variables

Positive Negative Sleep Reactions Reactions Disturbances

0–3 0–9 0–6________ ________ __________ Gender

Males 1.28 1.71 .48 Females .77 3.30 .94 t(228) 4.1*** 7.6*** 3.6***

Note. “Positive Reactions” includes “felt amused and entertained,” “laughed,” and “felt sexually turned on.” “Negative Reactions” includes “frequently hid my eyes or looked away,” “yelled and screamed,” “held onto my date,” “cried,” “heart beat fast,” “tried not to show how scared I was,” “was shaking,” “was generally very jumpy,” “felt disgusted,” and “my stomach felt funny.” “Sleep Disturbances” includes “fear of something specific,” “insomnia,” “fear of sleeping alone,” “needing to sleep with the lights on,” “serious or recurring nightmares,” and “wild imagination.” ***p < .001.

Table 5 presents participants’ratings of their own fear level while watching the movie and their remembered perceptions of how scared they wanted their date to think they were. Males were about as equally likely as females to circle the options “not scared” or “a little scared” and “somewhat scared” or “very scared,” whereas females were over four times more likely than males to circle “somewhat scared” or “very scared,” χ2(3) = 40.8, p < .01. In terms of how scared they wanted their dates to think they were, about a third of participants from each sex said they “didn’t care.” Of those who did care, men were twice as likely as women to remember that they wanted their date to think they were “not scared” or “a little scared,” whereas females were twice as likely as males to want their dates to think they were “somewhat scared” or “very scared,” χ2(3) = 19.4, p < .01. However, the difference between participants’ratings of their own fear and how fearful they wanted their date to think they were did not significantly differ for males (M = –.26) and females (M = –.29).

MEMORIES FOR SCARY M O V I E S 2 5 5

TABLE 5 R e p o rting of Own Fear Level and Pre f e rred Perception of One’s Fear by Date (%)

Question Asked About Fear Level “How scared did you want your

“How scared were you?” date to think you were?” ____________________ __________________

Males Females Males Females

“Not scared/ a little scared” 51 18 48 23

“Somewhat/ very scared” 46 80 19 40

“I don’t remember” 2 2 - -

“I didn’t care” - - 28 36

Note. Each scale had the same four choices: “not scared at all,” “a little scared,” “somewhat scared,” “very scared.” In addition, the “how scared were you?” question included the option “I don’t remember,” while the “how scared did you want your date to think you were?” question gave the option “I didn’t care.”

On a scale of 1 (completely unrealistic) to 7 (completely realistic), the realism of the movies viewed was rated a moderate 4.2. Overall, participants did not think the movies would scare them today (1 = not at all, 7= a lot), although males (M = 2.2) believed this more strongly than did females, M = 3.3, t(227) = 5.82, p < .001. On a 7-point scale from 1 (not at all) to 7 (very much), males (M = 5.3) reported liking this sort of film genre today more than females did, M = 4.7, t(227) = 2.86, p < .01. Accordingly, 47% of the males and 33% of the females reported that they later saw the movie again “several times,” 22% of the males and 20% of the females “once again,” and 28% of the males and 44% of the females “never” again, χ2(2) = 7.2, p < .05.

On a question regarding how the viewing had affected the feeling of emotional closeness to one’s date, there were no sex differences. Most of the participants (70% of males, 74% of females) checked the option that viewing the movie “did not affect” emotional closeness to one’s date. A minority (26% of males, 21% of females) checked the option “it made me feel emotionally closer,” whereas only 1%–2% checked “it made me feel more distant emotionally.”

2 5 6 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

Individual Difference Variables as Predictors of Viewing Reactions and Outcomes

The results pertaining to the personality measures are presented in Table 6. A series of multiple regressions was conducted on the data to assess the role that personality played in the experience and consequences of movie viewing. The correlations of relevant variables are presented in Table 7. College GPA, current age, age at viewing, and characteristics of the viewing situation were tested as possible covariates but were found to be unrelated to the outcome measures. The five dependent measures examined included the three previously described composite variables (Positive Reactions, Negative Reactions, and Sleep Disturbances), responses on a 7-point scale (Scared Today) asking “To what degree do you think that this movie would scare you today?” (1 = not at all, 7 = a lot ), and a composite measure “Liking,” which combined the two items “Did you ever see this movie again?” (choices of yes, several times; yes, once; and never) and a 7-point rating scale “Do you like this sort of movie today?” (1 = no, not at all, 7 = yes, very much). Significant predictors, with appropriate statistics, are summarized in Table 8 and are described in more detail below.

Sex and Gender Roles. The predictive value of biological sex as compared to sex-role behaviors as assessed by the PAQ was addressed through a series of hierarchical regressions. In the first set of regressions, one regression for each of the five dependent measures, the M, F, and MF subscores from the PAQ were entered in step one, and sex was entered as a second step. Although the MF scale was predictive of Negative Reactions to the scary movie, its value was overshadowed by sex when it was entered in step two. Biological sex, on the other hand, was predictive of all five criteria.

These results were confirmed via repeating the analyses using sex as step one and M, F, and MF scores from the PAQ as step two. When this was done, sex was a significant predictor of all five criteria, and the inclusion of PAQ scores did not significantly improve the equation.

The overall pattern of results suggests that males reported more Positive Reactions and more Liking for the genre. Females reported more Negative Reactions, Sleep Disturbances, and more likelihood by being Scared Today.

Empathy. The role of empathy was tested using the four subscale scores in a series of multiple regressions, one for each of the five criteria. Overall empathy predicted Negative Reactions, although none of the subscales by itself was significantly predictive. Sleep Disturbances were predicted by higher overall

MEMORIES FOR SCARY M O V I E S 2 5 7

2 5 8 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

MEMORIES FOR SCARY M O V I E S 2 5 9

empathy and Fantasy Empathy. Interestingly, Sleep Disturbances were also predicted by lower levels of Empathic Concern. Being Scared Today was predicted by higher levels of Empathic Concern and Personal Distress. Whereas the betas suggested that high empathy subscale scores were associated with

2 6 0 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

TABLE 8 Summary of Significant Individual Difference Predictors

Negative Reactions as predicted by Sex (Females report more

negative reactions) (ß = .49, p <.001; Total R2 = .24, p < .001) Empathy (R2 = .07, p < .05) Sensation Seeking (R2 = .08, p < .05)

Boredom Susceptibility (ß = –.20, p < .05)

Sleep Disturbances as predicted by Sex (Females report more sleep

disturbances) (ß =.26, p < .01; Total R2 = .08, p < .05) Empathy (R2 = .10, p < .01)

Fantasy Empathy (ß = .29, p < .001) Empathic Concern (ß =-.19, p < .05)

Being Scared Today as predicted by Sex (Females report being more

likely scared by movie today) (ß = .30, p < .001; Total R2 = .09, p < .01) Empathy (R2 = .12, p < .001).

Empathic Concern (ß = .20, p < .05) Personal Distress (ß = .20, p < .05).

Sex x Empathy interaction (ß = .17, p < .05; Total R2 = .15, p < .001) Females: High empathy women more likely Scared Today than low

empathy women. Males: Less likely than females to be Scared Today but no effect of

empathy

Liking as predicted by Sensation Seeking (R2 = .09, p < .05) Boredom Susceptibility (ß = –.24, p < .05)

Note. No other predictions significant at p < .05.

fewer Positive Reactions and less Liking (r = .15, p < .05), those equations were not significant.

Sensation Seeking. The effect of sensation seeking on the five criteria was also evaluated using multiple regression, with the Sensation Seeking subscales serving as predictors. Sensation seeking predicted Negative Reactions. Specifically, the higher the scores on Boredom Susceptibility, the fewer Negative Reactions were reported. Liking was predicted by low Boredom Susceptibility scores. None of the Sensation Seeking scales showed effects for Sleep Disturbances or being Scared Today.

Interactions Between Sex, Empathy, and Sensation Seeking. As seen in Tables 6 and 7, sex and the personality constructs of Empathy and Sensation Seeking are correlated. Therefore, multiple regressions were conducted with sex, total empathy, and total sensation seeking entered as a first step, with the Sex x Empathy and Sex x Sensation Seeking interactions entered on a second step.3

Two of the criteria were significantly predicted when all of the predictors were combined. Negative Reactions were predicted by Sex, such that females reported more negative reactions than males. Being Scared Today was predicted by the interaction of Sex and Empathy. Probes of the interaction (see Aiken & West, 1991) showed that high empathic females were more likely to be Scared Today than low empathic females. Males were less likely to report being Scared Today than females overall, and showed little change based on empathy.

Realism. One other variable that seemed to play a role is that of the perceived realism of the movie. As one can see in Table 7, Realism was related to Total Empathy, being Scared Today, and Liking. So, in an attempt to confirm findings of previous research (Hoekstra et al., 1999), the sex/empathy/sensation seeking analysis reported earlier was repeated adding realism as a predictor in a third step. Realism failed to make an independent contribution except in the case of Liking (= .20, p < .05), suggesting that the more realistic the movie is perceived to be, the more people like it; however, the inclusion of this step failed to bring the overall equation to significance, R2 = .07, p = .14.

Summary. Empathy and sensation seeking play a role in predicting emotion- related components of autobiographical memories about the experience of watching scary movies on a date, although not to the extent that one might expect. Clearly, the strongest predictor appears to be whether one is male or female, and this does not appear, at least according to our data, to be related to a measure of gender-role attitudes.

MEMORIES FOR SCARY M O V I E S 2 6 1

DISCUSSION

The Young Adult Dating Experience

Although previous research (Harrison & Cantor, 1999; Hoekstra et al., 1999) had established that children and teens are routinely exposed to violent movies that they remember as having some lasting effects, the setting of seeing the films was diverse in those studies. The present research has demonstrated that one common setting for seeing violent movies is on a date, an activity that almost every young adult experiences. These films are almost always viewed in the evening or at night, and they are equally likely to be selected by a male or a female.

On the face of it, watching violent movies with slashing and killing would appear to be inconsistent with the intimate context of a date, in contrast to, for example, a romantic movie. Yet such movies are a very common dating experience. This fact might merely reflect that such a large percentage of available movies are violent that seeing frightening movies is almost unavoidable. However, that is not an entirely satisfactory explanation. Although clearly many movies are violent, there are also many that are not. More likely, as noted by Zillmann and Weaver (1996), watching such movies actually enhances the date by allowing males to display fearlessness and protective competence and females to display fearfulness and protective need. Consistent with their model, the present data showed that females reported about twice as many Negative Reactions and Sleep Disturbances as males did, and males reported about twice as many Positive Reactions. Both sexes reported a moderate liking of the film, but males liked it even more. Females reported being more scared by the film than males did, although there was some tendency for males to want their dates to perceive them as less scared and females to want their dates to see them as more scared. Somewhat surprisingly, there were no sex differences in perceived emotional closeness resulting from the film, with about 75% of both sexes thinking the film had no effect. Most of the remainder thought it made them feel closer to their dates. However, this may be only a baseline effect that would occur on any date, regardless of the movie seen (or perhaps regardless of whatever activity the pair chose for their date). It may be that anything they did together would make them feel closer.

In terms of the regression analyses, it is noteworthy that none of the PAQ scales was as good a predictor of the various dependent variables as was biological sex. This finding might reflect that this measure, developed in the 1970s, is no longer as valid as it once was in assessing gender roles. Although this conclusion is certainly possible, another more interesting explanation seems

2 6 2 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

likely. It may be that social norms involved with dating are so powerful that they override gender attitudes, consistent with Eagly’s (1987) social role theory. According to this view, gender differences are the result of the roles that males and females play in social settings and the expectations built into those roles. Indeed, the social forces of the dating situation may be so strong as to influence one to act like a stereotypical man or woman, regardless of whether one actually holds traditional or egalitarian gender-role attitudes. These results support Zillmann and Weaver’s model and suggest that the laboratory effects obtained by Zillmann et al. (1986) were probably due to biological sex rather than gender- role attitudes of participants. This also confirms that their model generalizes from a controlled laboratory situation to memories of one’s own dating experiences.

Ordinarily, one would expect a sensitive attitudinal measure like the PAQ to be a better predictor of attitudes and reactions in a gender-relevant context like dating than a blunt, dichotomous variable like sex. The fact that this was not the case argues for a highly socially constrained situation in which one is strongly pressured to behave in highly stereotyped, almost ritualistic, ways. This is also consistent with many anecdotal reports from even very nontraditional males and females who find themselves acting more traditionally on a date.

Although various subscales of the empathy and sensation-seeking measures significantly predicted some Negative Reactions and Sleep Disturbances, as well as Liking and being Scared Today, these subscales were modest predictors at best. Empathic Concern was negatively correlated with Sleep Disturbances but positively correlated with being Scared Today, whereas Fantasy Empathy predicted sleep disturbances and Personal Distress predicted being scared (both positive correlations). The Boredom Susceptibility subscale was the only significant predictor from the Sensation Seeking scale, being negatively correlated with Liking and Negative Reactions. Finally, high-empathy females reported being significantly more Scared Today than low-empathy females, although this difference did not hold for males. Still, the variance accounted for by any of these predictors was very modest, especially when compared to sex, and more research is required to fully understand all relevant variables.

Limitations and Challenges

Although the reliabilities of the IRI, PAQ, and Sensation Seeking scales were acceptable (see Table 6), the low reliabilities of the additive composite measures of Negative and Positive Reactions and Sleep Disturbances were a concern. This may have been due in part to the evaluative ambiguity of some of the behaviors.

MEMORIES FOR SCARY M O V I E S 2 6 3

For example, although placed in the negative reaction group, “held on to my date” may have been seen as either positive or negative, depending on the person. Similarly, “felt sexually turned on” was included as positive but could have been troubling in the context of having such feelings during a violent movie. Although results from these three variables must be considered carefully, the large sex differences in almost all of the constituent items (Tables 2 and 3) remains a robust finding.

Another factor to consider was the possible variance of responses across the different movies. The participants were asked to remember a movie that had scared them, and they reported a great variety of films. Thus the commonality was the experience of being scared, not the type of film, although the large majority listed what are generally considered horror films. Given that 37% of the sample viewed either Scream or Scream 2, both popular theatrical films at the time, a second set of descriptive and regression analyses were run only on those seeing the Scream films. However, the reactions and predictions were essentially unchanged from the overall sample, so these analyses are not discussed further. The fact that the variances were not reduced in the subsample seeing the same movie, however, is reassuring in that it indicates that the complexities in the overall data were not likely due to noise from individual differences in the films viewed.

There are also some inherent limitations to the autobiographical memory methodology. The data are all retrospective in nature and thus subject to retrospective and possibly self-enhancing biases. For several reasons, however, these concerns are probably not critical in this case. First of all, because participants described a situation in which they admitted being fearful, any social desirability biases in reporting or self-aggrandizing biases in memory, as well as the general memory tendency for material to be forgotten over time, should have worked against the sort of reports obtained. Thus, it is quite likely that actual experienced fear at the time of viewing was greater than what was reported in the present studies. Therefore, results from the present study may have been, if anything, conservative compared to the actual effects.

Another concern might be the use of college-student samples. Although extension of this research to a broader sample would certainly be desirable to clarify and confirm the generalizability, there is every reason to expect that results would be similar. Indeed, negative effects might have been slightly mitigated in this relatively more educated sample. Factors such as overall media l i t e r a c y, parental supervision of viewing during childhood, and parental processing of negative experiences increase with level of education. Thus, as with the retrospective concern, this sample may indeed have underestimated the

2 6 4 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

effects. Also, teens and college students are typically a heavy movie-viewing demographic group. Thus, their experience and sophistication with scary films as young adults might be expected to produce greater desensitization to the very violent aspects, again potentially reducing the intensity of the reported negative effects. Finally, this particular sample was largely first-year students from a large public university with open admissions and a highly diverse student body in terms of ability and academic major. Thus, the usual concerns about selectivity of college samples, although still present, are less serious than would usually be the case.

Finally, the fact that the obtained negative reported experiences and effects are consistent with previous laboratory and field research showing a variety of negative consequences of viewing violent media also offers converging evidence for the effects obtained using a wide variety of methodologies.

Conclusions

The present study has extended the model of Zillmann and Weaver (1996) to the more ecologically valid setting of participants’ own dating experiences. The nearly universal experience of watching a violent movie on a date is a socially useful experience, but one that strongly constrains behavior in gender-prescribed ways, in line with Eagly’s (1987) social role theory. Males are pushed to display fearlessness and competence, females to display fearfulness and protective need.

Autobiographical memory offers a useful methodology to complement other techniques of investigating media effects. Future research would do well to study memories for viewing other genres of film and television. What are the uses and gratifications, for example, of watching a romantic movie on a date, and how does that differ from watching with a group of same-sexed friends? The present research has illuminated the social utility of watching frightening movies on a date, but that is by no means the only genre of film viewed in this context.

NOTES

1 Of these, one reported, “I don’t watch scary movies,” one reported, “I’ve never been scared,” and a third reported the movie Dr. Doolittle and made clear from his responses that he had not been scared by it.

2 Factor analysis resulted in multiple consistent clusters. However, because of abysmal reliability of subscales constructed of those factors, and recognizing the need to summarize findings beyond individual items, conceptually similar factors were combined to create subscales with the maximum amount of reliability.

MEMORIES FOR SCARY M O V I E S 2 6 5

3 Following the procedures from Aiken and West (1991), the variables were centered prior to construction of the interaction terms, and the centered variables were entered as predictors.

REFERENCES

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Brewer, W. F. (1995). What is recollective memory? In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory. Cambridge, UK: Cambridge University Press.

Cantor, J. (1991). Fright reactions to mass media productions. In J. Bryant & D. Zillmann (Eds.), Responding to the screen: Reception and reaction processes (pp. 169–197). Hillsdale, NJ: Lawrence Erlbaum Associates.

Cantor, J. (1994). Fright responses to mass media. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 213–245). Hillsdale, NJ: Lawrence Erlbaum Associates.

Cantor, J. (1996). Television and children’s fear. In T. M. Macbeth (Ed.), Tuning in to young viewers (pp. 87–115). Thousand Oaks, CA: Sage.

Cantor, J. (1998a). Children’s attraction to violent television programming. In J. Goldstein (Ed.), Why we watch: The attractions of violent entertainment (pp. 88–115). New York: Oxford University Press.

Cantor, J. (1998b). Mommy, I’m scared: How TV and movies frighten children and what we can do to protect them. San Diego: Harcourt Brace.

Conway, M. A., Bruce, D., & Sehulster, J. R. (1998). New directions in autobiographical memory research. Special edition of Applied Cognitive Psychology, 12(4).

Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44, 113–126.

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 and Social Psychology, 52, 126–133.

Donnerstein, E., & Smith, S. L. (1997). Impact of media violence on children, adolescents, and adults. In S. Kirschner & D. A. Kirschner (Eds.), Perspectives on psychology and the media (pp. 29–68) Washington, DC: American Psychological Association.

2 6 6 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG

Dubow, E. F., & Miller, L. S. (1996).Television violence viewing and aggressive behavior. In T. M. Macbeth (Ed.), Tuning in to young viewers (pp.117–147). Thousand Oaks, CA: Sage.

Eagly, A. (1987). Sex differences in social behavior: A social-role interpretation. Hillsdale NJ: Lawrence Erlbaum Associates.

Goldstein, J. (1998). Why we watch. In J. Goldstein (Ed.), Why we watch: The attractions of violent entertainment (pp. 212–226). New York: Oxford University Press.

Gunter, B. (1994). The question of media violence. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 163–211). Hillsdale, NJ: Lawrence Erlbaum Associates.

Harris, R. J. (1999). A cognitive psychology of mass communication (3rd ed.) Mahwah, NJ: Lawrence Erlbaum Associates.

Harrison, K., & Cantor, J. (1999). Tales from the screen: Enduring fright reactions to scary media. Media Psychology, 1, 97–116.

Hoekstra, S. J., Harris, R. J., & Helmick, A. L. (1999). Autobiographical memories about the experience of seeing frightening movies in childhood. Media Psychology, 1, 117–140.

Hoffner, C. (1997). Children’s emotional reactions to a scary film: The role of prior outcome information and coping style. Human Communication Research, 23, 323–341.

Johnston, D. D. (1995). Adolescents’ motivation for viewing graphic horror. Human Communication Research, 21, 522–552.

Lawrence, P. A., & Palmgreen, P. C. (1996). A uses and gratifications analysis of horror film preference. In J. B. Weaver, III & R. Tamborini (Eds.), Horror films: Current research on audience preferences and reactions (pp. 161–178) Mahwah, NJ: Lawrence Erlbaum Associates.

Lazarus, R. S. (1991). Emotion and adaptation. New York: Oxford University Press.

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.

Ross, B. M. (1991). Remembering the personal past: Descriptions of autobiographical memory New York: Oxford University Press.

Rubin, A. M.. (1994). Media uses and effects: Auses and gratifications perspective. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and re s e a rch (pp. 417–437). Hillsdale, NJ: Lawrence Erlbaum A s s o c i a t e s .

MEMORIES FOR SCARY M O V I E S 2 6 7

Rubin, D. C. (Ed.). (1986). Autobiographical memory. Cambridge, UK: Cambridge University Press.

Rubin, D. C. (Ed.). (1995). Remembering our past: Studies in autobiographical memory. Cambridge, UK: Cambridge University Press.

Spence, J. T., & Helmreich, R. L. (1978). Masculinity and femininity: Their psychological dimensions, correlates, and antecedents. Austin, T X : University of Texas Press.

Strasburger, V. C. (1995) Adolescents and the media: Medical and psychological impact. Thousand Oaks, CA: Sage.

Tamborini, R. (1996). A model of empathy and emotional reactions to horror. In J. B. Weaver III & R. Tamborini (Eds.), Horror films: Current research on audience pre f e rences and reactions. Mahwah, NJ: Lawrence Erlbaum Associates.

Zillmann, D. (1991). Empathy: Affect from being witness to the emotions of others. In J. Bryant & D. Zillmann (Eds.), Responding to the screen: Reception and reaction processes (pp. 135–167). Hillsdale, NJ: Lawrence Erlbaum Associates.

Zillmann, D. (1998). The psychology of the appeal of portrayals of violence. In J. Goldstein (Ed.), Why we watch: The attractions of violent entertainment (pp. 179–211). New York: Oxford University Press.

Zillmann, D., & Weaver, J. B. (1996). Gender-socialization theory of reactions to horror. In J. B. Weaver & R. Tamborini (Eds.), Horror films: Current research on audience preferences and reactions (pp. 81–101). Mahwah, 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 emotion. Journal of Personality and Social Psychology, 51, 586–594.

Zuckerman, M. (1994). Behavioral expressions and biosocial bases of sensation seeking. New York: Cambridge University Press.

2 6 8 HARRIS, HOEKSTRA, SCOTT, SANBORN, KARAFA & BRANDENBURG