Prior Exposure to Creatures From a Horror Film

Prior Exposure to Creatures From a Horror Film Live Versus Photographic Representations AUDFEY J. W E I S DOROTHY J. IMRICH BARBARA J. WnSON University of California, Santa Barbara

This experiment assesses the impact of two exposure strntegies on children's emotional and wgnitive reactions to afrightening movie scene. Childrenfrom twograde levels (kindergarten andfirst vs. second thmughfourth) received a desensitization treatment in which moa'eled exposure to a live earthworm was factorially varied with exposure to graphic photogzaphs of worms takenfrom a h m f i l m . Children then viewed afrightening scene involving worms tnkenfmm this m e f i l m . Results indimted that exposureto photographsincreased children's enjoyment of the movie segment and reducedf a r reactions to the scene. In wntmst, the live exposure strategy was effective in reducing fear reactions to the movie only among boys. However,live exposure did alter children's affective reactions to and judgments of w m themselves. Thefindings are discussed in terms of current theories of desemitiazfion and information processing.

raphic and frightening mass media presentations have become increasingly popular over the past decade. Friday the 13th Part VIII: JasonTakes Manhattan, Halloween V,and A Nightmare on Elm Street: The Dream Child are just a few examples of recent box office suc-

G

Audrey J. Weiss (MA, University of California, Santa Barbara, 1990)is a doctoral candidate in the Department of Communication, University of California, Santa Barbara. Her research focuseson the impactof mass media on children's cognitive and emotional reactions. Dorothy J. Imrz'ch is a doctoral student in the same department and is interested in the psychological effects of the mass media. Barbara I. Wilson (Ph.D., University of Wisconsin-Madison, 1985) is an associateprofessor in the Departmentof Communication, University of California, Santa Barbara. Her research focuses on developmentaldifferences in children's reactions to mass media. This study was supported by a grant to the third author from the College of Letters & Science, University of California, Santa Barbara. The authors would like to thank the following women for allowing children to be tested at their center or school:Terry Crump, Assistant Supervisor of Youth Activity at the Santa Barbara Recreation Department; Connie Staugaard, Director of the Salvation Army Day Camp; Principal Nancy Hill at Cleveland Elementary School; and Principal Jennie Dearmin at Adams Elementary School. The authors would also like to thank Jeff Levin for serving as one of the experimenters and Kimberly Schodtlerand Gabriel Gasca for coding the videotaped faaal expressions. Human CommunicationResearch, Vol. 20 No.1, September 1993 41-66 0 1993International CommunicationAssociation 41

42

HUMAN COMMUNICATION RESEARCH / September 1993

cesses. However, the accessibility of this type of content is not limited to movie theaters. Most theatrically-released horror films are now available for rent at family video outlets. In fact, the age of renters frequently goes monitored in video stores, thus allowing many children to obtain this material quite easily (Harris, 1989; Wartella, Heintz, Aidman, & Mazzarella, 1990). Furthermore, horror films currently featured on pay cable channels (e.g., Arachnophobia and Psycho N:The Beginning), syndicated television shows such as ”Monsters” and “Friday the 13th: The Series,“ and TV reruns of “Night Gallery” and ”Outer Limits” make frightening content accessible to viewers of all ages. The widespread availability of such content has led to a growing concern about the effects of these types of frightening presentations on young children. Recent studies show that a majority of preschool and elementary school children have experienced immediate fear reactions to mass media programming (Cantor & Reilly 1982;Cantor & Sparks, 1984; Cantor, Wilson, & Hoffner, 1986; B. Wilson, Hoffner, & Cantor, 1987). In addition, a substantialnumber of children have suffered fromnightmares, disturbing thoughts, and other long-term stress reactions to frightening programs (see Cantor, 1991, for a review). In a recent book on children’s fears, psychologist Sarafino (1986) cited mass media depictions, both fictional and nonfictional, as common causes of childhood phobias and traumas. The proliferation of this type of content and the difficulty of constantly monitoring children’s viewing render it nearly impossible to prevent young viewers from watching frightening programs. Moreover, many children report that they like scary programs, although these same children also frequently admit regretting having seen particular depictions (Sparks, 1986;B. Wilson et al., 1987).One possiblesolution is to let children discover their own strategies for dealing with scary programs. Yet research suggests that children do not always know how to cope with such presentations and that some strategies work better than others, particularly when considering the developmental level of the child (8.Wilson et al., 1987).Thus researchers and educators can play an important role in devising intervention methods to teach children techniques for coping with frightening media content. The issue of children’s responses to frightening mass media can be studied from a variety of perspectives. For example, researchers working from an interpretive perspective might examine the ways in which different children deconstruct the meaning of a particular frightening episode (see J. Anderson, 1981; Anderson & Avery, 1988; Delia, 1977). Other researchershave begun to document individual differencesin response to arousing media content (e.g., Sparks, Spirek, & Hodgson, 1992; Spirek & Sparks, 1993). This approach suggests that some children may be less susceptible than others to scary depictions. Although certainly not all

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

43

children will react uniformly to frightening portrayals, the studies cited above indicate that such depictions can pose difficulties for a majority of young viewers. In recognition of this widespread phenomenon, “Mr. Rogers’ Neighborhood” and ”Sesame Street” recently featured a series of episodesaimed at helping younger viewers cope with their media-related fears (Cantor, Sparks, & Hoffner, 1988; Leishman, 1981). One strategy that has been used successfully to treat children’sfears in clinical settings is desensitization (for reviews, see Firestone, Waters, & Goodman, 1978; Graziano, DeGiovanni, & Garcia, 1979). Desensitization techniques typically involve some form of gradual, repeated exposure to the fear stimulus under nonthreatening circumstances. Studies have found that such exposure strategies are effective in reducing young children’s fear of small animals (Tones, 1924; Lazarus, 1974), loud noises (O’Reilly, 1971; Tasto, 1969), and fire (Roberts & Gordon, 1979).Desensitization has also been shown to be effective in alleviating similar fears among older children (Murphy & Bootzin, 1973; Wish, Hasazi, & Jurgela, 1973). Several recent studies by B. Wilson (1987, 1989; B. Wilson & Cantor, 1987) have examined the use of desensitization techniques to reduce children’s fear reactions to the mass media. In one study, B. Wilson and Cantor (1987) used an educational film about snakes as the vehicle for desensitization. Children who received the visual exposure treatment watched 12 scenes of snakes, edited so that the movement and proximity of the reptiles gradually increased. All children then viewed a scary snakepit scene from Raiders of the Lost Ark. Results indicated that prior exposure to the educational clips of snakes tended to reduce self-reports of fear and heart rate during the frighteningmovie scene for both younger (5-7 years) and older (&9 years) children. However, these effects only approached significance. Furthermore, the visual exposure treatment had no impact on children’s judgments of the fear stimulus, such as whether most snakes can hurt people or are poisonous. In a second study, B. Wilson (1987) used a realistic facsimile of the fear stimulus for the exposure treatment. Children in the exposure condition were shown and allowed to play with a rubber replica of a tarantula. Then all children viewed a frightening scene involving tarantulas from the movie Kingdom of the Spiders. Prior exposure to the replica sigruficantly reduced negative reactions to tarantulas among both younger (5-7 years) and older (8-11 years) children. However, the exposure treatment did not affect self-reported degree of fear in response to the frightening scene. Furthermore, the treatment actually increased facialexpressionsof fear and disgust during the movie. And once again, the treatment had no impact on children’sjudgments about the dangerousness of tarantulas, a finding consistent with the earlier study (B. Wilson & Cantor, 1987). B. Wilson (1989) has argued that the limited effectiveness of the two exposure techniques described above might have been due to the nature

44

HUMAN COMMUNICATION RESEARCH / September 1993

of the presentations involving the fear stimulus. Recent theories of desensitization posit that gradual exposure serves to alter the perceptual (stimulus and response) features of a person’s fear image that is stored in memory (Foa & Kozak, 1986; Leventhal, 1984). According to this approach, the relative efficacyof a given exposure treatment depends on how closely the medium used to present the stimulus matches the stored fear image. Exposure treatments that closely resemble the stored fear image will evoke that image more efficiently and facilitate its alteration. Thus the filmed snakes and the rubber tarantula used in the above studies might not have corresponded closely enough with the children’s own images of real snakes and spiders to result in strong fear reduction. Current theories of desensitization also suggest why the two exposure treatments had no impact on children’s judgments of the fear stimulus. Leventhal(l984) and others (eg., Foa & Kozak, 1986) have argued that fear images are modified by inconsistentperceptualinformation about the fear object that is presented during treatment and that these iconic changes are relatively independent of higher-order cognitive processes. Such a perspective is supported by studies of habituation that indicate that even infants store perceptual information when exposed to novel stimuli (e.g., Bomstein & Benasich, 1986;Kaplan & Werner, 1986).Thus visual exposure techniques appear to operate at the perceptuallevel and do not necessarily alter more conceptual properties of the fear stimulus. Congruentwith these ideas, 8.Wilson (1989)in her third study devised two new treatmentsinvolving a live fear stimulus.In the passive exposure condition, children were allowed to observe a live, curly-tailed lizard, which presumably matched children’s own image of lizards more closely than did a filmed or a rubber stimulus.In the modeled exposure condition, children not only observed the live lizard but also watched the experimenter take the lid off the aquarium and touch the lizards back in a petting motion. This modeled exposure was designed to convey nonverbal information about the harmlessness of the fear object. After the treatments, all children viewed a scary lizard scene taken from the movie Frogs. Results indicated that both passive and modeled exposure reduced facial expressions of fear during the movie scene among younger (kindergarten-first grade) and older (second-fourth grade) children. However, only modeled exposure increased children’sself-reportedaffect toward lizards and their willingness to touch a live lizard. Furthermore, children in the modeled exposure conditionjudged lizards as less dangerous than other animals and more often wanted a lizard for a pet than did those in the passive or no exposure conditions. Similar modeling techniques have been used successfully in a multitude of nonmedia studies designed to reduce children’s fear of snakes (e.g., Weissbrod & Bryan, 1973), dogs (e.g., Hill, Liebert, & Mott, 1968), dental treatment (e.g., Melamed, Hawes, Heiby, & Glick, 1975),and sur-

Weiss et al. / PRIOR ExposuRETO FILM CREATURES

4!5

gery (Peterson, Schultheis, Ridley-Johnson, Miller, & Tracy, 1984). Such techniques have been found to be effective with both younger and older children (Ferguson, 1979; Komhaber & Schroeder, 1975; Melamed & Siegel, 1975). Furthermore, nonmedia research has also confirmed that modeled exposure typically is more effective than passive techniques (Johnson & Machen, 1973; Melamed, Yurcheson, Fleece, Hutcherson, & Hawes, 1978; White & Davis, 1974).For example, in one study, children who watched a model interacting with a dog subsequently were less frightened of dogs than were those who received only passive exposure to the animal (Bandura, Grusec, & Menlove, 1967). Bandura (1986) has argued that the primary function of observing a model’s behavior is to gain information about ways of responding to the fear stimulus and about the potential consequences of such interactions. In other words, the modeling helps a child to reconceptualize the fear stimulus as safe or nonthreatening. The results of B. Whon’s (1989) third study are consistent with this idea in that modeled exposure appeared to trigger the conceptual processing necessary to alter children’s judgments about lizards, whereas passive exposure did not. The three media studies together, then, suggest that exposure to a live stimulus might be more effective than exposure to a facsimile in reducing emotional reactions to the fear object itself. Moreover, modeled exposure can actually alter the child’s cognitions about the fear object. However, none of the treatments thus far have been very successful in reducing emotional reactions to a movie scene featuring the fear object. Modeled exposure and passive exposure have reduced facial fear during a movie scene (B. Wilson, 1989), but neither technique has affected children’s selfreports of emotional response to the movie. One likely explanation for this ineffectiveness is that none of the exposure treatments to date have prepared children adequately for the frightening depictions of the fear object as shown in a typical horror film. Presumably, prior exposure to a fear object produces a new perceptual image of the stimulus, one that probably corresponds closely to initial portrayals of the stimulusin a movie. Thus fearfulfacialexpressionsmight be reduced during early parts of the film segment. However, as the scene progresses and the stimulus becomes increasingly distorted and graphic, the child‘s newly formed mental image is likely to be challenged. By the end of the scene when self-reports of emotion are typically assessed, the exposure treatment might be rendered ineffective.In other words, it might be necessary to provide children with a representation of the fear object that more closely correspondsto the graphic portrayal of this stimulus on the screen. The purpose of the present study was to compare a relatively effective strategy, modeled exposure involving a live stimulus, with a new technique based on how the fear object actually looks in a graphic horror film.

46

HUMAN COMMUNICATIONRESEARCH / September 1993

These two desensitization techniques were factorially varied such that children were assigned to one of four conditions: exposure to a live stimulus in a modeling context, exposure to increasingly graphic photographs of the stimulus taken froma frighteningmovie, both live exposure and exposure to the photographs, or no exposure. The live exposure treatment is representative of modeling strategies used by parents and educators to combat children's fears. For example, one nursery school teacher in California combats preschoolers' fears of Halloween by showing children how to eat "ghost mush" and having them take naps with vampire dolls (S. Wilson, 1988).The photo exposure treatment also has a real-world analog in that a number of books and magazines are available today that deal with the making of horror films and contain photographs taken from frighteningmovie scenes (Horsting, 1986). Such photos could be presented gradually to children to prepare them for viewing an upcoming film. A recent Nao York Times article suggested a similar photo exposure strategy for preparing children to go to new places that are potentially frightening, such as showing children pictures of clowns before going to the circus (Kutner, 1991). After the exposure treatments, all children then viewed a brief clip from the same frightening movie used to create the desensitization photographs. It should be noted that watching a short excerpt is quite representative of the way in which children actually do watch television. Research indicates that children's attention to television is influenced by a variety of factors including the comprehensibility of a message, such that they normally attend to the screen for short periods of time (Anderson, Lorch, Field, & Sanders, 1981). Furthermore, watching a short clip is analogous to viewing previews or advertisements for movies, both of which frequently cause fear in children according to our discussions with many parents. Children from two age groups (kindergartenand first vs. second through fourth grade) participated in the study. These age groups were selected because they correspond to major developmental changes associatedwith Piagetian theory (Piaget, 1929) and with more recent models of information processing (Flavell, 1985). In particular, these theories indicate that around 7 or 8 years of age, children shift from a reliance on the perceptual or surface features of a stimulus to an increasing consideration of more conceptual properties in the same situation. Hence these two age groups reflect the two types of information processing involved in the two exposure treatments. However, none of the upcoming hypotheses in the present study predict developmental differences because prior research has shown that exposure techniques that convey conceptual information nonverbally are effectiveeven with young children (e.g., B. Wilson, 1989).

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

47

Based on the research reviewed above, the following hypotheses were formulated: H1: Both exposure to photographs of the stimulus and live exposure will alter children’s self-reported emotional reactions to the frightening movie scene, but exposure to photographs will be more effective overall.

This prediction is based on the idea that my type of prior exposure to the stimulus, either live or photographed, might help to desensitize children to the frighteningdepictions.However, the photo exposure technique was expected to have a stronger impact toward the end of the movie, when self-reports were assessed. By this point in time, photographs taken from the movie should prepare children more fully than would a live stimulus for the graphic, distorted images of the fear object typically presented during the latter portion of the frightening scene. Measures of emotion taken during earlier parts of the movie scene, however, might not show this superiority effect for the photo exposure component. When the fear object is portrayed more realistically, as is typically the case in the early scenes of frighteningfilms, both exposure to photographs and exposure to a live stimulus might help desensitizechildren to these initial depictions. Because facial expressions were assessed throughout the movie scene, a differential effect of the two components over time could be tested. A second hypothesis was posed: H 2 Both exposure to photographs and live exposure will influence children’s facial expressions during early portions of the movie segment, but exposure to the photographs will be more effective during the latter part of the movie.

In contrast to reactions to the movie, children’s emotional responses to the fear object itself should be altered by the treatment that most closely matches the child’s own perceptual image of the fear stimulus stored in memory. Based on this idea, a third hypothesis was formulated: H3: Exposure to a live stimulus will be effective in altering children’s affective reactions to the fear object, whereas exposure to photographs will have no such effect.

Live exposure that incorporates a modeling technique also provides conceptual information about the degree of threat associated with the fear object. Watching another person touch the fear object presumably helps to render the object as nonthreatening. On the other hand, exposure to photographs should be similar to any passive exposure technique that operates primarily at the perceptual level. Such a strategy does not yield

48

HUMAN COMMUNICATION RESEARCH / September 1993

any direct information about live interactions with the fear object. Consequently, a final hypothesis was constructed

H4 Exposure to a live stimulus in a modeling context will affect children’s judgments of the dangerousness of the fear stimulus, whereas exposure to photographs of the stimulus will have no such effect. METHOD

Participants The experiment involved 157children. Of these, 76 were in kindergarten and first grade (M age = 6.1 years, range = 4.8-7.7 years), and 81 were in second through fourth grade (M age = 8.6 years, range = 6.9-10.7years). The majority of the children were White (62%) or His- panic (28%),and the remainder were either Black (7%)or Asian (3%). Approximately half of the children were boys (n= 77)and half were girls (n = 80). Children attended either one of two recreation centers or one of two elementary schools,all in Santa Barbara, California.All children received written parental permission before participating.Within each grade level and within sex, children were randomly assigned to conditions. The ratio of boys to girls in each of the four experimental conditions was approximately equal for both grade levels. Design

In a 2 x 2 x 2 x 2 factorial design, children received one of four desensitization treatments based on two components: exposure to a live stimulus in a modeling context (no, yes), and exposure to photographs of the stimulus taken from a horror film (no, yes). The treatment conditions were varied with grade level (kindergartenand first vs. second through fourth grade) and with sex of subject. All children then viewed a frightening scene from the same horror film that was used to create the desensitization photographs. Procedure Before any testingbegan, the children met with the three experimenters (twofemales and one male) for a brief explanation of the research project. All childrenwere tested individuallyin two quiet rooms. In the first room, the experimenter presented the child with one of the various versions of the desensitization treatment. The experimenter then escorted the child

Weiss et al. / PRIOR ExposURE TO FILM CREATURES

49

into the second room to watch a videotape during which his or her facial expressionswere recorded. During selected periods of black screen edited onto the videotape, the experimenter asked questions about the child's immediate reactions to the program just viewed. The viewing session lasted approximately 20 minutes. The experimenter then escorted the child back to the first room and asked a series of questions about the program and about the fear stimulus. In addition, a behavioral measure of approach-avoidance toward the stimulus was taken. The postviewing session lasted approximately 15 minutes. Before returning to the classroom, the experimenter asked each child not to discuss the study with the other children until all testing was completed. Desensitization Treatments

All desensitizationtreatments were administered befoIe children viewed the videotape. The first component of desensitization involved modeled exposure to live earthworms. In the live exposure-only condition, the experimenter introduced the treatment as follows: Before we start, I'd like to show you something that I brought with me. Do you know what worms are? They are mushy animals that crawl under ground. I brought some worms today so you can see them. Let me show you.

The experimenter then placed a clear glass bowl in front of the child. The bowl contained dirt and approximately one dozen live earthworms. The experimenter then said, Do you see the worms? They crawl around in the dirt. Some worms are very small and skinny. And some are big and fat. These are fat worms. If you reach into the bowl, you can pick up a worm and hold it .. ,like this.

The experimenterthen reached into the bowl, picked up a worm, and held it out for the child to see. The experimenter=turned the worm to the bowl and said, Ok, why don't you look at the worms for a few minutes while I get things ready for the TV shows.

Children were allowed to observe the worms for approximately 2 minutes. The second component of desensitization involved exposure to three photographs of worms that were taken from the horror movie Squirm? Two of the photographs were from the actual movie segment that the children would be viewing. In the photo exposureonly condition, the experimenter said: Before we start, I'd like to show you something that I brought with me. Do you know what worms are? They are mushy animals that crawl under

50

HUMAN COMMUNICATION RESEARCH / September 1993

ground. I brought some pictures today of how worms sometimes look on TV.Let me show you. The experimenter presented the three pictures successively, in a gradual fashion, each time noting that this is “another picture of worms on TV.“ The first picture showed several worms crawling on the ground, the second picture depicted a pile of hundreds of worms in a bathtub, and the third picture displayed a close-up shot of worms attacking a man’s face. After arranging the three pictures in a row in front of the child, the experimenter said, Ok, why don’t you look at the pictures for a few minutes while I get things ready for the TV shows. Children were allowed to observe the pictures for approximately2 minutes. Children in the combined exposure condition were exposed both to the live worms and to the three photographs. To control for possible order effects, half of the children were first exposed to the live worms and then to the pictures, whereas the other half viewed the pictures first and then the live worms. The experimenter presented the first component in the combined exposure condition as described above and allowed the child to view the live worms/pictures for approximately 2 minutes. The experimenter then removed the first desensitizationcomponent and introduced the second component as follows: I also brought [some pictures of how worms sometimes look on TV/some live worms for you to see]. Let me show you.

The experimenter then presented the second component as described above and allowed the child to view the pictures/live worms for approximately 2 minutes. Finally, children in the control condition were not exposed to any worms, either live or photographed from the movie, prior to viewing. The children simply waited quietly with the experimenter for approximately the same amount of time as the treatments lasted. To insure that all treatments, including the control, were approximately equal in terms of length of time the experimenter used an hourglass as a timer to mark the exposure periods. Materials The videotape viewed by the children consisted of two key segments. The first segment (2 minutes, 23 seconds) was taken from a nonarousing, educationalprogram about weaving cloth. This segment was used to relax children and to acclimatethem to the experimentalsetting. The second segment (2 minutes, 51 seconds) was a frightening scene involving worms, taken from the 1976 PG-rated movie Squirm. During the segment, a man

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

51

and woman are shown fishing in a boat on a lake. As they talk, the camera zooms in on a container of worms in the boat. The container then tips over, and the worms spill out onto the bottom of the boat. As the man tries to move closer to the woman, he falls over in the boat and the worm begin to attack his face. The boat eventually capsizes, and the scene ends with the panicked man rushing out of the water. Dependent Measures

Self.Reports of Emotional Response to the Program Two self-report measures of emotional response were assessed. Immediately following the Squirm segment, the experimenter asked, “How scared did you feel during this last program? Did you feel not at all scared (0),a little &ifscared (l),pretty scared (2), v e y scared (3), or very v e y scared (4)?” During the voicing of this question the experimenter displayed a laminated board that depicted five illustrated faces, exhibiting increasing amounts of fear, with the corresponding label below each face in increasing letter size. The second measure of emotional response involved degree of affect toward the program. The experimenter asked, “How did you feel about the program? Did you like the program or did you dislike the program?”* Depending on the selected alternative, the experimenter then asked, ”How much did you likddislike the program?Did you (dis)like it a little bit, (disjlike it pretty much, (disjlike it very much, or (disjlike if v e y v e y much?“ The experimenter showed a board that displayed the appropriate response options for either ”like” or ”dislike” in increasing letter size.

Facial Expressions Children’s facial expressions provided a nonverbal indicator of emotional reactions during the Squirm segment. Facial expressions were recorded by a small color camera placed next to the television set. The camera contained a stopwatch function that allowed tenths of a second to be superimposed on the videotaped recordings. Most of the children appeared to be unaware of the camera equipment. Facial expressions were coded using the Affex Coding Scheme (Izard, Dougherty, & Hembree, 1983). The Affex system involves the identification of appearance changes in the face that correspond with 10 emotion expressions. An expression can occur in the entire face or in the upper or lower region of the face only, permitting the identification of blends of expressions. Coders coded the onset and offset of all facial expressions exhibited during the Squirm segment. Two expressions, fear and enjoy-

52

HUMAN COMMUNICATION RESEARCH / September 1993

ment, were of primary interest because they coincided with the self-report measures of emotional response to the movie and because theoretically they should be influenced by the desensitization treatments. The frequency of each expression was defined as the number of times the expression occurred either in the full face or as part of a blend?

SelfReported Affect Toward Worms One self-report measure was used to assess children’s affect toward worms. Immediately after children viewed the Squirm segment, the experimenter asked whether they liked or disliked worms. Depending on the selected alternative, the experimenter then asked, “HOWmuch do you like/dislike worms?Do you (didlike thema little bit, (dis)like them pretty much, (disjlike them very much, or (disj1ik-ethem very very much?” The experimenter again displayed the “like” or “dislike” options on a board while voicing this question.

Judgment of Worms Three measures were used during the postviewing session to assess children’s conceptions of worms. The first two measures involved judgments about the dangerousness of worms. First, the experimenter asked, ”If your friend saw a worm on the sidewalk and was really scared, what would you tell him/her?” Responses subsequently were coded into one of six categories: those indicating that the friend should not be scared or worried (e.g., ”don’t be scared), those indicating that the worm was not dangerous (e.g., “it won’t bite you”), those indicating that the friend should kill the worm or stay away from it (e.g., “step on it”), those indicating that the child would protect the friend (e.g., “I’d probably say, ’I’ll protect you’ ”), miscellaneous responses (e.g., “that it was there”), or “don’t know” responses. Responses that indicated either that the friend should not be scared or that the worm was not dangerous subsequently were collapsed because they indicated that the child understood that the worm was not a threat to self or other. The remaining four categorieswere combined to compare nonthreatening responses to all other responses. Second, the experimenter asked, ’What about worms and people? Do you think worms cannot hurt people at all (0),can hurt people a little bit (l), can hurt people pretty much (2), can hurt people very much (3), or can hurt people very vety much (4)?” During the voicing of this question, the experimenter displayed a board that depicted the alternatives in increasing letter size. The remaining judgment measure involved children’s desire to have a live earthworm versus a live snail as a pet. The experimenter showed children a picture of a worm and a picture of a snail. Both pictures were approximately the same size and were laminated on cardboard. The

Weiss et al. / PRIOR ExposURE TO FILM CREATURES

53

experimenter then asked, "Which would you want more for a pet, a worm or a snail?"4

Behavioral Approach-Avoidance At the end of the session, a measure of the behavioral tendency to approach or avoid worms was taken. The experimenter presented each child with the bowl of live earthworms and asked, "Would you like to touch one?" The experimenter recorded whether or not the child actually reached into the bowl and touched one of the live worms. Data Analyses For all categorical data, multidimensional contingency tables were created using the live exposure component of desensitization, the photo exposure component, sex, and grade level as factors. Log-linear analyses were conducted on the frequencies of responses for such data. This type of analysisprovides a parallel to the traditionalANOVAfor nominal-level data (Marascuilo & Levin, 1983).For all analyses, both the marginal and partial tests of association were performed. Because both types of tests resulted in the same statistical decision, only the partial likelihood ratio chi-squares (G') are reported.Post hoc comparisonswere conducted using the chi-square analog to the Scheff6procedure (p < .05). For interval-level data, unweighted-means analyses of variance were performed using the same four factors.Painvise comparisonswere made using the Scheff6 procedure < .05). Table 1 presents a summary of all main effects for the live exposure component and photo exposure component of desensitization for all dependent measures.

Reliabilities Two judges who were blind to the children's grade level and experimental conditioncoded the open-ended responses to the question of what to tell a scared friend. Disagreements were solved by a thirdjudge. Based on Scott's (1955) pi, intercoder reliability was 96.7% for this question. For the facial expressions, two undergraduate coders achieved 80% reliability on the Affex training materials prior to any coding. Each coder then coded the onset and offset of facial expressions for half of the children. Coderswere blind to the children's grade level and experimental condition. In addition, both coders coded expressions for 10 randomly

54

HUMAN COMMUNICATION RESEARCH / September 1993

TABLE 1

Effects for Each of the Two Components of Desensitization on All of the Dependent Measures ~~

Dependent Measure Self-reported emotion toward program Fear in response to program Liking of program Facial expressions Part 1 Fear Enjoyment Part 2 Fear Enjoyment Self-reported affect toward worms Judgmentof worms Suggestions to a friend Harmfulness of worms Want pet worm more than pet snail Behavioral response

~

Type of Desensitization Live Exposure Photo Exposure

-

* *

*

-

*

-

* *

*

*

-

NOTE: Table represents main effects for the live exposure component and the photo exposure component of desensitization.Asterisk indicates the presence of a main effect for the designated component of desensitization.

selected children. Intercoder reliability was 94.1%(range = 85.2%-99.5%) based on the proportion of time that coders were in agreement or half agreement (in the case of blends) on the expressions identified. Self-Reportsof Emotional Response to the Program The data for the self-report measure of fear (how scared the child felt) were skewed, so responses to this question were collapsed into two categories: those indicating no fear versus those indicating some degree of fear. A log-linear analysis was performed on the frequency of fearful feelings in response to the frighteningmovie scene. Results revealed a sigruficant main effect for the photo exposure component of desensitization, G2(1,N = 157) = 4.33, p < .05, V* = .02? Children who had seen the photographs were sigruficantly less likely to report feelingscared by the Squimz segment (48%) than were those who had not seen the photographs (64%). This analysis also revealed a significantmain effect for grade level, GZ(1, N = 157) = 14.95, p < .001,V" = .07.Younger children reported a signifi-

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

55

M

e

. I

UY

m a4 L P

s C

t

s

rs

. I

No Live Exposure

Live Exposure

Figure 1 Percentage of Boys and Girls Expressing Fear in Response to the Movie Segment as a Function of the Live Exposure Component of Desensitization

cantly lower proportion of fearful feelings in response to the movie Scene (41%) than did older children (70%). In addition, there was a two-way interaction between sex and the live exposure component of desensitization, G2(1,N = 157) = 5.57, p c -05, V* = .03. In particular, pairwise comparisons revealed that boys who were exposed to the live worms were sigruficantly less likely to report fearful feelings (%YO,) than were boys who hadnot seen the live worms (59%,,, see Figure 1).In contrast, exposure to the live worms had no impact on girls’ reports of fear during the frighteningsegment (liveexposure: 70”/ob;no live exposure:6O%J. Finally, there was a main effect for sex, G2(1, N = 157) = 6.17, p = .01, V‘ = .03. However, this main effect was due primarily to the significant interaction between sex and live exposure. No other effects were sigruficant in this analysis. For the measure of affect toward the program, the data were also somewhat skewed, so responses to this question were similarly combined into two categories: those indicating liking the program versus those indicating not liking the program. A log-linear analysis of the percentage of liking responses resulted in a significant main effect for the photo exposure component, G2(1,N = 157) = 5.99, p = .01, V* = .03. Children who were exposed to the photos reported liking the frighteningprogram more (82%) than did those who did not see the photos (65%).

56

HUMAN COMMUNICATION RESFARCH / September 1993

Facial Expressions In accord with Hypothesis 2, the Squirm segment was partitioned into two parts for the analyses of facial expressions. During Part 1 of the segment (1minute, 58 seconds), the worms were shown first in a container and then as they spilled out onto the bottom of the boat and slithered around. During Part 2 (53 seconds), the worms were depicted in a more graphic and distorted fashion as they attacked the man’s face. As with the data for self-reports of emotional response, the data for the frequency of facial expressions were also skewed because of a few children who were quite expressive relative to the majority. Thus facial expressions were transformed into categorical data and a log-hear analysis was conducted on the percentage of children who exhibited an expression versus those who did not. For Parts 1and 2 of the segment, separate log-hear analyses were performed on the two primary facial expressions, fear and enjoyment. For Part 1,the analysis of fear expressions revealed a sigruficant main effect for photo exposure, G2(1,N = 157) = 4.42, p < .05, V* = .02. Children who had seen the photographs were less likely to exhibit facialfear during Part 1 of the Squirm segment (6%)than were those who had not seen the photos (17%).No other effects emerged in this analysis of facial fear. Unlike the facial fear data, the analysis of enjoyment expressions during the first part of the movie scene revealed no sigruficant effects. For Part 2 of the Squirm segment, the analysis of fear expressions resulted in a sigruficant interaction between grade and sex, G’(1, N = 157) = 8.43, p < .01, V* = .04.Pairwise comparisons revealed that older girls were sigruficantly more likely to exhibit facial fear (12”/0) than were younger girls (O%,). There was no difference in facial fear between younger boys (110/4) and older boys (5%&). The analysis also revealed an interaction between grade and the live exposure component of desensitization, G’(1, N = 157) = 3.83, p = .05, V+ = .02. However, none of the simple pairwise comparisons were sigruficant. The analysis of the enjoyment expression during this second time period resulted in a number of findings. First, there was a main effect for the photo exposure component of desensitization,G2(1,N = 157) = 3.66, p = .05, V* = .02. Children who had seen the photographs were more likely to exhibit facial enjoyment (190/,)during Part 2 of the Squirm scene than were those who had not seen the photos (9%). Second, this analysis revealed a significant main effect for sex, G’(1, N = 157) = 8.77, p < .01, V* = .04. Boys were significantlymore likely to exhibit facial enjoyment (22%) than were girls (6%). Finally, the analysis of the enjoyment expression during the second time period also revealed a sigruficantthree-way interaction between the photo exposure component, the live exposure component, and sex, G’(1,

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

57

N = 157) = 7.29, p < .01, V* = .03. However, none of the meaningful comparisons with the control p u p within each sex were significant. Self-Reported Affect Toward Worms An analysis of variance was conducted on the ratings of how much children liked worms. The analysis resulted in a highly sigruficant main effect for the live exposure component of desensitization,F(1,141) = 10.91, p = .001, qz = .07. Children who were exposed to live earthworms in a modeling context reported that they liked worms more (M = 4.7) than did those who were not exposed to worms (M = 3.4). The only other significant finding was a main effect for sex, F(1,141) = 14.98, p < .001, q2= .lo. Boys liked worms (M = 4.9) more than did girls (M = 3.2).

Judgment of Worms Two measures were used to assess children’sjudgments of the dangerousness of worms. For the open-ended question regarding what to tell a scared friend, a log-linear analysis was performed on the frequency of responses that indicated that the worm was not a threat. The analysis resulted in a highly significant main effect for the live exposure component of desensitization, G’(1, N = 157) = 16.22, p < .OOl, V*= .07. Children who had been exposed to the live earthworms were sigruficantly more likely to indicate that a worm posed no threat (71%)than were those who had not seen the live worms (41%). The only other significant finding in the analysis was a main effect for grade level, G2(1, N = 157) = 14.70, p < .001, V*= .07. Older children were sigruficantly more likely to indicate that a worm posed no threat (69%)than were younger children (41%). For the second question, the data for children’s ratings of how much worms can hurt people were skewed, so responses to thisquestion were collapsed into two categories: those indicating that worms cannot hurt people versus those indicating that worms can hurt people. A log-linear analysis on the frequency of hurt responses revealed a sigruficant main effect for the live exposure component of desensitization, G2(1,N = 156)= 4.22, p < .05, V* = -02.6Children who had been exposed to the live worms were significantly less likely to report that worms can hurt people (26%) than were those who had not seen the live worms (41%). The remainingjudgment measure concerned the idea of having a worm for a pet. In particular, children were asked which they would like more for a pet: a worm or a snail. The log-linear analysis of the frequency of ”worm” responses revealed a sigruficant main effect for the live exposure component, G2(l,N = 157)= 7.68, p < .01, V* = .a. Children who had been exposed to the live worms were more likely to want a pet worm over a

58

HUMAN COMMUNICATION RESEARCH / September 1993

pet snail (69%)than were those who had not seen the live worms (48%). The only other sigruficantfinding in this analysis was a main effect for the photo exposure component of desensitization, G2(1,N = 157) = 5.96, p < .05, V* = .03.Unexpectedly, children who had seen the photographs were less likely to want a pet worm over a pet snail (49%)than were those who had not seen the photos (68%). Behavioral Response The log-linear analysis of the frequency of children who agreed to touch a live worm revealed a significantmain effect for live exposure, G2(1,N = 151) = 7.53, p c .01, V‘ = .04?Children who had been exposed to the live worms in a modeling context were sigruficantly more likely to touch a worm (66%)than were those who had not been exposed to the live worms

(44%). The analysis also revealed a significant interaction between the photo exposure component and grade level, G2(1,N = 151) = 4.68, p < .05, V* = .02. However, none of the meaningful comparisons between the photo exposure and no exposure conditions were significant within each grade level. DISCUSSION The results of this study generally indicate that the photo exposure strategy was more effective than live exposure in modifying children’s emotional reactions to the frighteningprogram. In contrast, live exposure was more successfulin altering affectiveresponses to the fear object itself. Furthermore, live exposure was the only strategy to modify children’s judgments about the dangerousness of the fear object. In terms of specific predictions, the first hypothesis posited that both exposure to photographs and exposure to a live stimulus would alter children’s self-reports of emotional response to the frightening program but that the photo exposure strategywould be more effective overall.This prediction was supported in two ways. First, both desensitization strategies reduced self-reported fear responses to the movie scene. In particular, children who had seen the photographs were significantly less likely to report feeling scared by the program than were those who had not seen the photos, The live exposure treatment strategy also reduced self-reported fear but not for all children. Boys who had seen the live earthworms were less likely to report feeling scared by the program than were boys who had not seen the live worms, but no such difference emerged for the girls. Robust sex differences that were observed for two other measures may help explain this pattern. Specifically, boys were more likely than

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

59

girls to exhibit facialexpressionsof enjoyment during the latter part of the movie and were more likely to report liking worms. Similarsex differences have been observed in previous media studiesinvolving snakes (B. Wilson & Cantor, 1987), spiders (B. Wilson, 1987), and lizards (B. Wilson, 1989) and are consistent with research that indicates that girls tend to be more frightened of small animals and insects than are boys (Lapouse & Monk, 1959; Ollendick, 1983). Consequently, a live exposure strategy might be less successful with girls who, because of their general dislike of reptiles, insects, and snakes, might be resistant to looking at such creatures and to appreciating them in movie portrayals. In addition to self-reportedfear responses, the first hypothesis was also supported by self-reports of affect toward the program. However, in this case only the photo exposure strategy produced an effect. In particular, those children who had seen the desensitization photographs reportedly liked the movie scene more than did those who had not seen the photos. Taken together, the findingsindicate that both live exposure and exposure to photographsinfluenced self-reportedemotionalresponses to the movie scene, but the photo exposure strategy had a far more extensive effect because it altered both fear and liking of the movie and did so for all children. Notably, this experiment is the first in a series of media studies on desensitization to demonstrate that an exposure strategy cun significantly affect self-reported emotional reactions to a movie scene. As mentioned previously, passive techniques, such as exposure to a rubber replica or to educational clips of the fear stimulus, have not been effective in altering self-reports of fear to a frightening movie (B. Wilson, 1987; B. Wilson & Cantor, 1987). The present findings for the photo exposure strategy suggest that during passive exposure treatments, representations of the fear stimulus should closelycorrespond to the depictions in an upcoming film. In this study,children saw photographs that were actually taken from the movie segment they were about to view. Thus the treatment images strongly resembled the fear object as depicted in the movie scene, In line with this reasoning, prior exposure to motion picture previews or movie clips might produce even more potent desensitization effects than still photos because the representations would be completelyisomorphic with the upcoming film.Future research should test exposure strategies that increasingly simulate the portrayal of fear objects in movies. The second hypothesis predicted that both exposure strategies would alter children’s facial expressions during early portions of the movie segmentbut that exposureto photographs would be more effectiveduring the latter part of the movie. This hypothesis was only partidy supported. As expected, the photo exposure strategy did influence facial expressions during both the early and later portions of the movie. Children who had ilm were less likely to been exposed to the photographs taken from the f

60

HUMANCOMMUNICATION RESEARCH / September 1993

exhibit facial fear during Part 1of the segment than were those who had not seen the photos. In addition, children who had seen the photographs were more likely to express facial enjoyment during Part 2 of the movie scene than were those who had not seen the photos. Thus the photo exposure strategy was so efficacious that it not only reduced fear early in the scene but also actually helped children enjoy the most intense and graphic part of the segment. In contrast to our prediction, however, the live exposure strategy was not effective in altering children’s facial expressions during the early, less intense part of the movie scene. One possible explanation for thisineffectiveness is that the live earthworms might not have represented a close enough match to the worms even as they were initially depicted in the first half of the segment. That is, the live exposure component fostered an image of worms as inactive, docile creatures in a glass container of dirt. This newly formed image might not have prepared children sufficiently for the depictions in the early part of the movie scene when worms were portrayed not yet as harmful but as very actively squirming around in a boat. The ineffectivenessof the live exposure strategyin influencing initial facial expressions once again underscores the importance of a strong perceptualcorrespondencebetween the desensitization treatment and the upcoming fear stimulus. The third hypothesis predicted that live exposure to worms would modify children’s emotional reactions to the fear stimulus itself but that exposure to photographs would not. This prediction was strongly supported in two ways. First, only the live exposure component affected children’s liking of worms. In particular, those who were exposed to live earthwormsreported a sigruficantlyhigher degree of affect toward worms than did those who were not exposed to live worms. Second, only the live exposure strategy affected the behavioral measuxe of fear of worms. Children who watched amodel pickup a live earthworm were more likely themselves to touch the worms than were those who did not see this modeled demonstration.The live exposure treatment presumably helped to establish a perceptual or visual image of real earthworms,which should have been useful when reacting emotionally to worms themselves. The photo exposure strategy, on the other hand, only helped foster a perceptual image of how worms can be distorted in a movie. No images of docile or realistic earthworms were provided during this treatment, thereby limiting its effectivenesson emotional responses to the fear stimulus itself. The fourth hypothesis predicted that the live exposure strategy would alter children’s judgments of the dangerousness of the fear stimulus but that exposure to photographs would not. This prediction received substantial support from all three judgment measures of worms. Children who had been exposed to live earthworms were more likely than those who had not seen the worms to (a) tell a scared friend that worms pose

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

61

no threat, (b) rate worms as less harmful to people, and (c) indicate that they would like a pet worm more than a pet snail. None of these effects were observed for the photo exposure component.The differentialimpact of the two strategies is consistent with the idea that modeled exposure provides conceptual information about the nonthreatening nature of the stimulus (Bandura, 1986), whereas photographic representations do not necessarily provide this type of information. It should be noted that the third judgment measure, comparinga worm and a snail for a pet, did reveal a main effect for the photo exposure strategy but in the reverse direction than would be predicted by desensitization. Specifically, children who had seen the photographs were less likely to choose a pet worm over a pet snail compared to those who did not see the photos. There are at least two possible explanations for this finding. One is that perhaps photo exposure did provide conceptual information about earthworms, but in this case the graphic photos fostered a more rather than less dangerousimpression.However, this account seems unlikely given that none of the otherjudgment measures revealed an effect for photo exposure. A more plausible explanation concerns the fact that, unlike the other judgment measures, during the voicing of thisparticular question a picture of a worm (and a snail) was shown to facilitate responses. Children in the photo exposure conditionsmighthave associated this new photograph of an earthworm with the earlier, perceptually displeasing desensitization photos and thus avoided selecting the picture of the worm for the preferred pet. Overall, the two exposure strategies were successful for both younger and older children.This pattern is consistent with previous media studies involvingnonverbal exposure treatments (B. Wilson, 1987,1989)and with nonmedia studies that indicate that desensitization techniques can be effective with very young children (e.g., Bender, 1962; Wallick, 1979). However, there were general age differencesfor some of the dependent measures. In terms of emotional reactions to the program, older children were more likely than younger children to report feeling scared and to tell a scared friend that worms pose no threat. One explanation for these findings is that older children might have had more experience with worms in real life. For instance, teachers often instruct children about the relative harmlessness of many small animals and insects, some of which are kept as classroom pets. Thus the distorted image of worms presented in the movie scene might have upset the older children more because these images contrasted so sharply with thisage p u p ’ s more deeply ingrained mental image of real worms. One other point is worth noting with regard to the overdl findings. It might be argued that children who received both the live exposure plus the photo exposure components should have evidenced the strongest desensitization effects in this study. Indeed, these children actually re-

62

HUMAN COMMUNICATION RESEARCH / September 1993

ceived twice as much desensitization as did those children in either condition alone. Alternatively, it could be argued that the two components combined should have confused children rather than enhanced desensitization because in one strategy worms are portrayed as docile and in the other worms are seen as dangerous. However, the persistent main effects for either one or the other of the components on all of the dependent measures suggest that the two strategies operated relatively independently, with exposure to the photos influencingreactions to the movie and exposure to live worms affecting responses to worms themselves. This idea is further supported by the fact that no meaningful interactionswere observed between the two desensitization techniques. That is, the combined condition appeared to neither help nor confuse children compared to either treatment alone. It is quite likely that the children in the combined exposure condition simply reacted to each of the two components separately without even noticing the discrepant information, especially since our experimental instructions did not explicitly connect the two types of images.Future researchmightexaminedifferent methodsfor simultaneously showing children how a real stimulus looks and how it often is distorted when portrayed in frightening films. In conclusion, this study replicates and extends previous research on desensitization strategies in the context of frightening media presentations. Wenow have evidencefrom this study and an earlier one (B. Wilson, 1989)that modeled exposure to a live stimulus can have some impad on children’semotional reactions to aprogram (althoughlimitedonly to boys in the current study), and that this type of treatment influences affective reactions to and judgments of the fear stimulus itself. Unlike previous studies, however, this experiment examined a new exposure strategy. Presenting children with increasinglygraphic photographs of a fear stimulus taken from a frightening movie scene alters children’s emotional responses to that scene. However, this technique does not appear to influence children’s affective reactions to and judgments of the fear stimulus itself. The findings have a number of important implications. Theoretically, the two desensitization strategies employed in this study illustrate the differential impact of perceptual versus conceptual cues presented during exposure. The photo exposure technique altered Children’s perceptual or iconic representations of the fear stimulus, thereby affecting emotional responses to the movie scene. In contrast, the live exposure strategy altered children’s mental conceptions of the fear stimulus and subsequently influenced judgments of the fear object itself. Indeed, current models of information processing distinguish between bottom-up processing, which is driven by the perceptual features of a stimulus, and

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

63

top-down processing, which is governed by more conceptual aspects of the stimulus (Daehler & Bukatko, 1985). This research also suggests several practical implications for parents and educators.Parents could prepare children in advance for a frightening media presentation by showing them photographs from the program, such as those found on the back of video rental boxes or in movie magazines. Additionally, parents or teachers could introduce children to benign creatures by interacting with the animals under nonthreatening circumstances. For example, an adult could demonstrate for a child how to put ants into an ant farm. This type of exposure might discourage the child from drawing generalizations about ants from a single frightening media depiction, such as a movie about killer ants. Future research on desensitization strategiescan proceed in a number of directions. Studies should further explore the effects of prior exposure to photographic representations of frightening stimuli. Perhaps verbal explanations can be used in conjunction with photographic strategies to help children comprehend the nonthreateningnature of creaturesthat are inaccurately portrayed in mass media. However, such explanationsmust be designed to be consistent with the cognitive abilities of younger children (e.g.,B. Wilson & Weiss, 1991).Studiesshould also attempt to identlfy the possible long-term benefits of live versus photographic exposure techniques. Conceptual-level strategies, such as live modeling, could potentially extend beyond a single desensitization episode and influence children‘s reactions in future encounters with the fear stimulus. Research on desensitization ultimately will increase our understanding of children’s cognitive and affective processing of mass media.

NOTES 1. All of the photographs from the movie that were shown to the children were created by photographing freeze-frameshots of the television screen.The standard-sizephotographs were then mounted on cardboard and laminated. 2. The response options were counterbalanced for this question and all subsequent questions involving two response alternatives. 3. The two facial expressions of fear and enjoyment were also measured in terms of duration, defined as the total number of seconds the expression occurred either in the full face or as part of a blend. Analyses of duration, however, revealed no significant findings because of the high variabilityof this measure across subjects. 4. The two response options, and consequently the order of the pictures, were counterbalanced for this question. 5. Estimates of variance accounted for in all log-linear analyses arebased on the log-linear analog to Cramer’s (V*) measure of association (Marascuilo & Levin, 1983). 6. The reduced number of subjects in this analysis is due to missing data for 1child. 7. The reduced number of subjects in this analysis is due to missing data for 6 children.

64

HUMAN COMMUNICATION RESEARCH / September 1993

REFERENCES Anderson, D. R, Lo&, E. P., Field, D. E., &Sanders, J. (1981). The effects of TV program comprehensibility on preschool children’s visual attention to television. Child Dmlopmt,52,151-157. Anderson, J. A. (1981).Reseaxh on children and television: A critique. Journalof Broadcasting, 25,395-400. Anderson, J. A., & Avery, R. K. (1988). The concept of effects: Recognizing our personal judgments. Journal of Broadcasting 6 Elecfronic Media, 32,359-366. Bandura, A. (1986).Socialfoundationsofthought and action:A social cognitive the0y .Englewood Cliffs, NJ: PrenticeHall. Bandura, A., Grusec, J. E., &Menlove, F. L. (1967).Vicarious extinctionof avoidance behavior. Journalof Personality and Social Psychology, 5,16-23. gentler, P.M. (1962).An infant’s phobia treated with reciprocal inhibition therapy. Journal of Child Psychology and Psychiatry, 3,185-189. Bornstein, M. H., & Benasich, A. A. (1986).Infant habituation: Assessments of individual differences and short-term reliability at five months. Child Dmelopmmt, 57,87-99. Cantor, J. (1991). Fright responses to mass media productions. In J. Bryant & D. Zillmann (Eds.), Responding to thesmm: Reception and readion processes (pp. 169-197).Hillsdale, NJ: Lawrence Erlbaum. Cantor, J., & Reilly, S. (1982).Adolescents’ fright reactions to television and films. Journal of Communication,32(1), 87-99. Cantor, J., & Sparks, G. G. (1984).Children’s fear responses to mass media: Testing some Piagetian predictions. Journal of Communication,34(2), 90-103. Cantor, J., Sparks, G. G., & Hofher, C. (1988). Calming children’s television fears: Mr. Rogers vs. The Incredible Hulk. Journal of Broadcasting b Electrunic Media, 32,271-288. Cantor, J., Wilson, B. J., & Hoffner, C. (1986). Emotional responses to a televised nuclear holocaust film. CommunicationResearch, 13,257-277. Daehler, M. W., & Bukatko, D. (1985).Cognitive development. New York Alfred A. Knopf. Delia, J. G. (1977).Constructivism and the study of human communication. Quarterly Journal ofspeech, 63,66-83. Ferguson, B. F. (1979). Preparing young children for hospitalization: A comparison of two methods. Pediatrics, 64,656-664. Firestone, P., Waters, B. C., &Goodman, J. T. (1978).Desensitization in children and adolescents: A review. Journal of Clinical Child Psychology, 7,142-148. Flavell, J. H. (1985).Cognitive development (2nded.). Englewood Cliffs, NJ: Prentice-Hall. Foa, E. B., & Kozak, M. J. (1986).Emotional processing in fear: Exposure to corrective information. Psychological Bulletin, 99,2045. Graziano, A. M., DeGiovanni, I. S., &Garcia, K. A. (1979).Behavioral treatment of children’s fears:A review. Psychological Bulletin, 86,804-830. Harris, R J. (1989).A cognitive psychology of mass communication. Hillsdale, NJ: Lawrence Erlbaum. Hill, J. H., Liebert, R. M., & Mott, D.E.W. (1968).Vicarious extinction of avoidance behavior through films. An initial test. Psychological Reports, 22,192. Horsting,J. (1986).S t e p k King at the m o z h . New York Starlog Press. Izard,C. E.,Dougherty,L.M.,&Hembree,E.A. (1983).Asystemforidmti~ingaffectexpressions by holistic judgments (A8ex-J.Newark University of Delaware, hStruCtiOMl Resources Center. Johnson, R., & Machen, J. B. (1973).Behavior modification techniques and maternal anxiety. Iournul of Dentisty for Children, 40,272-276. Jones, M. C. (1924).The elimination of children’s fear. Journal ofExpm’menta1Psychology, 7, 382-390.

Weiss et al. / PRIOR EXPOSURE TO FILM CREATURES

65

Kaplan, P. S., & Werner, J. S. (1986).Habituation, response to novelty, and dishabituation in human infants: Tests of a dual-processtheory of visual attention. Journalof Expm'mental Child Psychology, 42,199-217. Komhaber, R C., & Schroeder, H. E. (1975). Importance of model similarityon extinction of avoidancebehavior in children. Journal of Consulting and Clinical Psychology, 43,601607. Kutner, L.(1991, January31). Assurances of safety calm jitters. The New York Times, p. C8. Lapouse, R., & Monk, M. A. (1959).Fears and worries in a representativesample of children. Americrm Journal of Orthopsychiaty, 29,223-248. Lazarus, A. A. (1974). The elimination of children's phobias by deconditioning. In H. J. Eysenck (Ed.), Experiments in behavior therapy (pp. 467-474). New York Pergamon. Leishman, K. (1981, January 10).When is television too scary for children?TV Guide, pp. 4-8. Leventhal, H. (1984). A perceptual motor theory of emotion. In K. R Scherer & P. Ekman (Eds.), Approaches to emotion (pp. 271-292). Hillsdale, NJ: Lawrence Erlbaum. Marascuilo, L. A., & Levin, J. R (1983).Multiuariate statistics in the social sciences. Monterey, CA: Bmks/Cole. Melamed, B. G., Hawes, R. R., Heiby, E., & Glick, J. (1975).Use of filmed modeling to reduce uncooperative behavior of children during dental treatment. Journal of Dental Research, 54,797-801. Melamed, B. G., & Siegel, L. J. (1975).Reduction of anxiety in children facing hospitalization and surgery by use of filmed modehg. Journal of Consulting and Clinical Psychology, 43, 511-521. Melamed, B. G., Yurcheson, R., Fleece, E. L., Hutcherson, S., & Hawes, R. (1978). Effects of filmed modeling on the reduction of anxiety-related behaviors in individuals varying in level of previous experience in the stress situation. fournal of Consulting and Clinical P ~ y ~ h o l46,1357-1367. ~g~, Murphy, C. M., & Bootzin, R R (1973). Active and passive participation in the contact desensitizationof snake fear in children. Behavior Therapy, 4,203-211. Ollendick, T. H. (1983). Reliability and validity of the Revised Fear Survey Schedule for Children (FSSC-R). Behavior Research and Therapy, 21,685-692. (YReilly, I? P. (1971).Desensitization of firebell phobia. Journal ofschool Psychology, 9,5557. Peterson, L., Schultheis, K., Ridley-Johnson,R., Miller, D. J., &Tracy,K. (1984).Comparison of three modeling procedures on the presurgical and postsurgical reactions of children. Behavior Therapy, 15,197-203. Piaget, J. (1929).The child's conception of the world. London: Kegan Paul. Roberts, M. C., &Gordon,S. B. (1979).Reducing childhoodnightmaressubsequent to a bum trauma. Child Behavior Therapy, I, 373-381. Sarafino, E. P.(1986).Thefears of childhood: A guide to rewgnizing and reducing fearful states in children. New York Human Sciences Press. Scott, W. A. (1955).Reliability of content analysis: The case of nominal scale coding. Public Opinion Quarterly,19,321-325. Sparks, G. G. (1986).Developmental differences in children's reports of fear induced by the mass media. Child Study Journal, 1 6 , 5 5 4 . Sparks, G. G., Spirek, M. M., & Hodgson, K. (1992, May). Individual diferences in amusability: Implicationsfor understanding immediate and lingering emotional reactions tofnghtening mass media. Paper presented at the annual meeting of the International Communication Association, Miami, FL. Spirek, M. M., & Sparks, G. G. (1993, May). The impad of children's coping style on emotional reactions f oafigkfening movie. Paper presented at the annual meeting of the International Communication Association, Washington, DC. Tasto, D. L. (1969).Systematic desensitization, muscle relaxation and visual imagery in the counterconditioning of a four-year-old phobic child. Behavim Research and Therapy, 7, 409-411.

66

HUMAN COMMUNICATION RESEARCH / September 1993

Wallick, M. M. (1979).Desensitization therapy with a fearful two-year-old. Amm’can Journal of Psychiatry, 236,1325-1326. Wartella, E., Heintz, K. E., Aidman, A. J., & Mazzarella, S. R. (1990).Television and beyond: Children’s video media in one community. Communication Reseurch, 17,454. Weissbrod, C. S., & Bryan, J. H. (1973). Filmed heahnent as an effective fear-reduction technique. Journal of Abnormal Child Psychology, 1,196-201. White,. W. C., & Davis, M. T. (1974). Vicarious extinction of phobic behavior in early childhood. Journal of Abnormal Child Psychology, 2,25-32. Wilson, B. J. (1987). Reducing children’s emotional reactions to mass media through rehearsed explanation and exposure to a replica of a fear object. H u m n Communication Research, 14,326. Wilson, B.J. (1989).Desensitizing children’s emotional reactions to the mass media. Communication Research, 16,728-745. Wilson, B. J., & Cantor, J. (1987). Reducing fear reactions to mass media: Effects of visual exposure and verbal explanation. In M. McLaughlin (Ed.), Communication yearbook 10 (pp. 553-573). Newbury Park, CA. Sage. Wilson, B. J., Hoffner, C., & Cantor, J, (1987). Children’s perceptions of the effectiveness of techniques to reduce fear from mass media. Journal of Applied Developmental Psychology, 8,39-52. Wilson, B. J., & Weiss, A. J. (1991). The effects of two rvality explanations on children’s reactions to a frightening movie scene. Communication Monogruphs, 58,307-326. Wilson, S. W. (1988, October 23). Play does trick for scared children at Halloween. Santa Barbara N m s Press, pp. D1, D3. Wish, P.A., Hasazi, J. E., & Jurgela, A. R. (1973). Automated direct deconditioning of a childhood phobia. Journal of Behavior Therapy and Expm’mental Psychiatry, 4,279-283.