34.6 Attitudes and Instructional Media - The Literature

In the last 15 years, two attempts have been made to organize the literature related to attitude change and instructional technology. In the late 1970s and early 1980s, Simonson reported on a series of reviews of the literature that culminated in a set of guidelines for designing mediated messages for obtaining attitudinal outcomes. Several studies were reported during the following years that attempted to validate Simonson's (1979) guidelines.

In 1993, Bednar and Levie proposed a series of attitude change principles. These principles facilitated the design of instruction to change attitudes and for the processes used for structuring lessons that targeted affective goals.

Simonson's research will be summarized first, then Bednar and Levie's principles will be reviewed. Finally, there will be a discussion of other research literature that relates to the use of media for persuasion.

34.6.1 Designing Instruction for Attitudinal Outcomes--The Iowa State Studies

In 1977, the first of a number of reviews and research studies dealing with media-attitude research was published by Simonson. A three-part approach was followed. First, literature about media and attitudes was located, reviewed, and synthesized (Simonson & Burch, 1977; Simonson, Thies & Burch, 1979a; Simonson, 1979; Simonson, 1980). Next, Simonson proposed a series of guidelines for designing instruction for attitudinal outcomes (Simonson, 1979, 1983, 1994). Finally, a series of research studies were conducted that evaluated various aspects of Simonson's guidelin6s (Simonson, 1985; Simonson, Aegerter, Berry, Moock & Stone, 1987; Treimer & Simonson, 1988; Dimond & Simonson, 1988). This research agenda will be reviewed.

Several publications by Simonson reviewing the media and attitude literature were published during the 3-year period between 1977 and 1980. The purpose of these reviews was to summarize the status of the research in this area and to attempt to synthesize conclusions about the results of this research. Simonson identified 211 research studies that experimentally examined some aspect of the relationship between attitudes, attitude change, and instructional media. Published or abstracted in AV Communications Review were 138 articles, and 73 articles were found that were published elsewhere (Simonson, Thies & Burch, 1979; Simonson, 1979a).

Simonson (1979a) arrived at several general conclusions after reviewing this literature. Five observations, with each having impact on those preceding it, were offered.

Observation #1: Mediated instruction does contribute to desired attitudinal outcomes in learners, especially when the instruction is designed specifically to produce certain attitudes or attitude changes.

Observation #2: The state of the art of media-attitude research is such that no specific guidelines for producing attitudinal outcomes can be generated. 'Ibis is no theory of "media-produced attitude change."

Observation #3: Procedures most likely to produce desired attitudinal outcomes toward content as a result of instruction delivered by media include:

• Use of follow-up activities (e.g., discussions) and open-ended questions after the mediated instruction
• Maximum use of realistic types of media with as many nondistracting/noncontradictory visual cues (e.g., color, motion) as possible (e.g., the story film)
• Creation of an aroused state in the learner through direct participation, intermittent distractions to reorient the learner from previously held attitudinal positions, or dramatic presentations that involve the viewer emotionally and intellectually in the content shown

Observation #4: Procedures least likely to produce desired attitudinal outcomes toward content as a result of instruction delivered by media include:

• Varying only the channel through which the instruction is delivered (e.g., a videotaped replay of a live lecture probably would not produce more favorable viewer attitudes toward the lecture topic, and might even produce more negative attitudes than would the live lecture)
• Including distractions in the form of extraneous information or cues in the mediated instruction (e.g., a videotaped presentation on the need for better study habits probably would not change viewers' attitudes toward the use of the public library as a source of information even though this was included as part of the videotape script; too much irrelevant information inhibits attitude change)
• Presenting unrealistic, highly symbolic information using visual media (e.g., a slide set which included many frames containing written information probably would not produce positive attitudes toward the topic presented)
• Presenting problem/conflict situations but providing no mechanism for conflict alleviation or problem solving, either as part of the mediated messages or in follow-up discussions (e.g., a film on the problem of soil erosion probably would not change farmer's attitudes unless the film showed solutions to the problem of soil erosion)

Observation #5. Media-attitude experimentation is not currently a high priority for researchers in media or related areas.

Simonson (1979a) stated that these five observations were opinions, but that any careful review of the 221 studies summarized would produce the same or similar conclusions. Simonson (1980) stated that if mediated instruction was broadly defined to include the entire learning process of which television, film, or still pictures were a part, then mediated instruction did seem to contribute to attitude formation and change. When only the media were evaluated, then conclusions were much less conclusive. Simonson stated that only one, broad, general conclusion about the relationship between media and attitudes was apparent and that this conclusion was an obvious one. Instructional media are primarily carriers of information and play their greatest role in the attitude change process as delivery vehicles. Characteristics of media such as flexibility of use, accessibility of information, and ability to encode ideas were more important than any inherent communication-related characteristics of a medium, which probably were of secondary importance to any development of attitudes or attitude changes (Simonson, 1980).

Simonson, Thies, and Burch (1979) also identified four trends in media-attitude research. These trends were listed as phases that characterized the research about media and attitudes.

Phase #L Liking. A number of early research studies evaluated the media-attitude link in the simplest form by attempting to determine if the learner liked the lesson delivered by media or the medium itself. Usually, researchers reported on the results of an evaluation of learners. They were asked questions like, "Did you like this lesson?" or "Do you like learning from film?"

Phase #2. Change in liking. The logical second level of attitude evaluation conducted by media researchers was the study of change in liking as a result of media exposure. Usually pre- or postsurveys of learner attitude were conducted in order to determine if there was a positive or negative change.

Phase #3. Attitude comparison. A slightly more sophisticated design was the use of two treatments to determine the impact of attitude change procedures. Often, one treatment was a mediated one and the other was traditional, teacher-centered instruction. These comparison studies suffered from the same problems as media comparison studies of achievement: Results from many studies produced contradictory findings.

Phase #4. Media-attitude interaction. In the late 1970s, researchers began to design experimental studies using an aptitude-treatment-interaction approach (Allen, 1975). This approach made it possible to determine with greater specificity the relationship between attitudes and media. While relatively few studies were found that used this approach, the ones that were found did provide considerable information for the development of guidelines for designing persuasive mediated instruction (Simonson, 1979).

Next, a number of methodological problems common to the media-attitude literature were summarized by Simonson. These concerns related directly to the problems encountered by those who subsequently attempted to synthesize this area of research.

First, attitudes were poorly defined. Often researchers did not operationally define the attitude constructs they were investigating. Rather, statements such as "attitudes were measured and students liked learning about chemistry from films" were reported. Measurement of attitudes was also considered inadequate. Most measures were locally prepared, and most researchers did not report validity or reliability information about their attitude tests.

Most studies reviewed by Simonson were considered to be quasi-experimental or experimental (Campbell & Stanley, 1963). Often, however, attitude testing was only of peripheral importance to the main purposes of the research study. Many times, attitudes were evaluated after the fact~ as follow-ups to an experiment. Clearly stated attitude hypotheses or research questions were rarely reported.

Finally, it was rare that follow-ups of the results of attitude change treatments were reported. Many critics of attitude research considered then, as they do now, that attitudes are transitory and attitude changes short-lived. Social psychology literature tends to refute this criticism. Few media-attitude researchers reported studying the long-term consequences of their persuasive efforts.

Simonson (I 979a) concluded that the research on media and attitudes was not significantly different from media research in general. However, Simonson did feel that there were important implications to the research that made possible the proposal of general guidelines pertaining to the design of mediated instruction for attitudinal outcomes. Simonson offered a series of six guidelines (1979, 1983, 1984), which are included next.

Several versions of six guidelines for designing mediated instruction for attitudinal outcomes were proposed by Simonson' (1979, 1983, 1984). In all cases, they were Supported by research literature and were offered with two purposes in mind. First, the guidelines were generated for instructional developers to apply during the instructional design process (1979). They were also offered to researchers so they could test the effectiveness of the guidelines.

Simonson's six guidelines for designing instruction of attitudinal outcomes are:

Guideline #1: Learners react favorably to mediated instruction that is realistic, relevant to them, and technically stimulating.

Guideline #2: Learners are persuaded, and react favorably, when mediated instruction includes the presentation of new information about the topic.

Guideline #3: Learners are positively affected when persuasive messages are presented in as credible a manner as possible.

Guideline #4.- Learners who are involved in the planning, production, or delivery of mediated instruction are likely to react favorably to the instructional activity and to the message delivered.

Guideline #5: Learners who participate in postinstruction discussions and critiques are likely to develop favorable attitudes toward delivery method and content.

Guideline #6.- Learners who experience a purposeful emotional involvement or arousal during instruction are likely to change their attitudes in the direction advocated in a mediated message.

Simonson (1979) concluded his discussion of these six guidelines by repeating the position, taken several times previously, that media are primarily carriers of information. There was no "best medium" found for producing attitude outcomes. However, there apparently was a best approach for the maximizing of the likelihood of desirable attitudes being fostered in learners in a specific situation. The guidelines were thought to be useful for developing specific attitude change messages.

Apparently, the six guidelines received a lukewarm reception from the profession. While no rebuttals were published, neither were there many research studies published that cited the guidelines as the basis for either a research plan or an instructional design process. One exception to this were the studies conducted at Iowa State University during the 1980s that attempted to investigate one or more of the guidelines. Six of these studies will be discussed next.

34.6.3 Media and Attitude Change: Six Studies

Simonson's (1979) guidelines were derived from previous research dealing with media and attitudes. In order to evaluate them, a number of research studies were completed and reported. The first four were 'published as a group (Simonson, Aegerter, Berry, Kloock & Stone, 1987). One was published in the journal Teaching and Learning Technologies (Simonson, 1985) and replicated and presented at a convention of the Association for Educational Communications and Technology (Dimond & Simonson, 1988), and the last study was published in the Journal of Social Psychology (Treimer & Simonson, 1988).

The first four studies were part of a research program that investigated whether instructional media could be used to deliver persuasive messages effectively. The studies attempted to provide evidence about the following questions:

1. Is there a hierarchy of media types related to effectiveness in delivering persuasive messages? In other words, are media that deliver messages realistically, as defined by Dale (1946), more effective than media that depict messages less realistically?
2. Does learner aptitude interact with media type when attitude change is the goal of the message? Specifically, do the learner characteristics of field dependence/ field independence and hemisphericity interact with media type when persuasive messages are delivered?

Study #1. This study attempted to determine if a motion picture was more effective than a nearly identical slide with audiotape presentation at changing attitudes of viewers toward the need for greater soil conservation efforts. This study used a posttest only, control group design (Campbell & Stanley, 1963). There were three randomly assigned groups of subjects, two experimental and one control. The two experimental treatments were based on a 23-minute film entitled We Are of the Soil. This motion picture was designed to introduce soil conservation practices, such as conservation tillage, and to convince viewers that these practices were important. It was selected by a panel of experts as a film that was technically excellent (Cook, 1979), and as one that seemed to have been produced more to change attitudes than to provide information.

Subjects in the first treatment viewed the motion picture. Students in the second treatment watched a 2 X 2 slide presentation and listened to an accompanying audiotape. To produce this slide presentation, each scene in the film was analyzed, and the most important still picture from each scene was made into a slide. The film's narration was duplicated on an audiotape. The slides were projected using a dissolve unit and two carousel slide projectors, a treatment that was reported by Cook (1979) to be comparable in technical quality to the motion picture from which it was derived. The motion picture was considered to be the medium that would deliver the message the most realistically (Dale, 1946).

Subjects. Students enrolled in an undergraduate teacher education course were assigned randomly to one of the three treatment groups. Before treatments were administered, all subjects were given the Group Embedded Figures Test (GEFT) (Witkin, Oltman & Raskin, 197). The GEFT was used to determine whether students tended to be either field-dependent (FD) or field-independent (FI) learners. This learner characteristic was examined because it was felt that the impact of the mediated treatments might be different for subjects who were either FD or Fl.

FD and FI are considered to be pervasive, stable cognitive styles that influence a person's perception of messages (McLeod, McCormack, Carpenter & Skvarcius, 1978). FD learners are influenced more by their surroundings than are F1 learners, who are influenced more by internal factors. FD individuals seem to be more socially oriented and are more affected by praise and criticism from their peers. FD persons tend to take a passive, spectator role in learning. FI learners, on the other hand, seem more adept at taking a message apart and at understanding its component parts, and tend to be more active learners who often have a strong self-concept. People are not totally field dependent or field independent, rather they have tendencies one way or the other.

Subjects were given the GEFT, then were categorized as being either FD or Fl, depending on their score on the GEFT. Students who had scores within one raw score of the average of all scores were not included in treatments. They were excused from the experiments because the GEFT did not satisfactorily identify them as tending to be either FD or Fl.

Design. This study employed a 2 X 3 factorial design. There were three treatment groups. One group watched the motion picture, one group watched the slide presentation, and the third was a control group. Within each group there were two levels of each treatment: those students considered to be FD and those considered to be FI.

After treatments were viewed, subjects were given the Soil Conservation Attitude Test (SCAT). The SCAT was developed by Cook (1979) and revised by Kloock (198 1). It contained 24 statements to which subjects reacted using a five-response Likert-type scale. The SCAT was reported to have a reliability estimate of .85 (Kloock, 198 1).

Results of Study #1. Descriptive statistics are reported in Table 34-2. There was a statistically significant difference in attitude reported that was attributable to the experimental treatments. The average scores of students in all four experimental treatment cells indicated a more positive attitude toward the importance of soil conservation than did the scores of control subjects. Because one of the main goals of this study was to determine if any of the experimental groups of subjects reacted to treatments significantly differently from the others, a Duncan's test (Ferguson, 197 1) was used to identify where significant differences occurred among cells. It was found that the subjects in the motion picture treatment who were identified as being field independent had more positive attitudes than did subjects in any of the other five treatment groupings.

Study #2. This study was a modified replication of study 1, with the same general purpose. However, three changes were made. First, the topic of the experimental treatments was changed. A film entitled The Right Approach was selected by a jury of media specialists as an excellent persuasive film. Its topic was the employment of the handicapped. A slide presentation, with accompanying audiotape, was produced from the most relevant scenes of the film in a manner similar to the slide treatment produced for study 1. The two treatments (motion picture and slide with audiotape) were judged by experts to be generally of comparable technical quality.

Since the topic of the treatments changed, the test of the dependent variable also had to be changed. A standardized text of attitude toward disabled persons was found in the Mental Measurements Yearbook (Buros, 1978). The Attitudes Toward Disabled Persons (ATDP) was reported to have a reliability estimate of .76 (Yucker, Block & Young, 1970).

The second change for study 2 was to use fifth- and sixth-grade students as subjects. They ranged in age from 10 to 13. Last, a follow-up test was given to a small sub-sample of subjects 3 weeks after treatment to determine if attitude changes produced by the treatments persisted.

The posttest-only design for this study had two independent variables, field dependence/field independence and treatment. The 2 X 3 factorial design had three treatments (motion picture, slide with audiotape, and control), and two levels of the cognitive style, field dependence/independence.

Results of Study #2. Results of descriptive tests are reported in Table 34-3. There was a statistically significant difference in attitude attributable to treatments and to the learner cognitive style field dependence/independence. After treatments, the subjects who had viewed the motion picture generally had more positive attitudes toward disabled persons than did subjects who watched the slide/audiotape presentation. Average attitude scores of subjects in both of the experimental treatments were significantly more positive than were the average scores of control subjects.

Average scores of several treatment cells deserve note. Three of the experimental treatment cells, Film/FD, Film/Fl, and Slide/Fl, had approximately equal attitude scores, while the fourth treatment group, Slide/FD, had significantly less positive attitudes toward the disabled than did students in any of the other three experimental cells.

This study added a follow-up testing session. Three weeks after treatments were administered, a smaller number of the subjects were randomly selected for retesting. There were no significant differences found, even though the trends of scores were similar to those obtained from the original administration of the attitude test. There seemed to be a regression effect. It was also apparent that field-independent subjects generally remained more positive toward disabled persons than did field-dependent subjects. However, because such a small number of subjects were included in this retesting, it is not possible to draw generalizable conclusions from the data.

Study #3. This experiment could also be considered a modified replication of study 1. For this experiment, there were two major changes made to the design of study 1. First, junior and senior high school students were used as subjects. These students attended school in a small town in an agricultural state in the Midwest and ranged in age from 13 to 18. The second change in this study was the examination of the independent variable hemisphericity instead of field dependence/independence.

Researchers have reported that in spite of a great deal of overlap of function, the two hemispheres of the brain organize and encode information in two different ways. Generally, the left hemisphere is more logical, convergent, and analytical- It is responsible for language and processes information sequentially. The right hemisphere is more holistic, intuitive, spatial, and divergent (Ornstein, 1977).

It also has been determined that individuals tend to have a dominant hemisphere. That is, one hemisphere tends to take priority when information is processed. It has been proposed that this hemispheric dominance is related to effective learning. In other words, how a person's brain perceived data determines in part how much is learned.

In order to assign subjects to treatments, the Conjugate Lateral Eye Movement (CLEM) Test was used to identify a person's dominant hemisphere (Day, 1964). The CLEM is an individually administered test that requires observation of a subject's eye movement after reflective questions are asked of them. The CLEM Test has a reliability of .78 (r = .78).

Subjects were tested using the CLEM and were assigned to treatment groups. Individuals who did not have a clearly dominant hemisphere, as indicated by the CLEM, were excused from the experiment. Specifically, subjects who did not move their eyes in a predictable pattern when they were asked reflective questions were not included in data analyses.

The first treatment group viewed the persuasive film We Are of the Soil. The second group watched the slides with accompanying audiotape prepared from this motion picture. The third group was a control. After treatments were completed, the SCAT was administered. This study used a 2 X 3, posttest only, control group design.

Results of Study #3. Descriptive statistics are reported in Table 34-4. There were no statistically significant differences found, although the trends of the mean scores were interesting. Left-brain-dominant subjects generally were more positive than were right-brain subjects in alltreatment categories. In order to examine the data more completely, a post-hoc analysis of SCAT scores for subjects in grades 9 through 12 was conducted. While results were not significant, it was found that subjects in the senior high school grades who were in experimental treatments had more positive attitudes when compared to control subjects in the same grade. In other words, the difference between control group subjects' and experimental group subjects' attitude scores were greater in the higher grades than they were in the lower grades.

Study #4. This experiment took a slightly different approach than studies 1, 2, or 3. At its foundation was the principle reported by Simonson (1984) and Rogers (1973) that the use of fear may be an effective technique for attitude change, especially if cures to the problem or probabilities of exposure to a fear-provoking event are included in the message. In other words, a persuasive message that showed the dire consequences of not following some course of action-such as stopping smoking, or wearing seat belts-could be made more effective if cures for the problem or techniques for how to change behavior were included in the message.

Study #4 used a 2 X 3, posttest only, control group design. As before, field dependence/independence was the second factor in the design. This learner characteristic was hypothesized as possibly being related to the impact of a fear-provoking message, especially when information to reduce the tension produced in viewers as a consequence of the fear was included in one treatment and not in the other.

The college students who participated in the experiment were tested using the-GEFF and assigned to one of the three treatment groups, just as they were in study 1.

Experimental treatments were based on a film entitled The Feminine Mistake, a 23-minute, antismoking motion picture sponsored by the American Cancer Society. This film was selected by a group of media specialists because as obtained from of its high technical quality. Permission w the copyright holder to produce two 15-minute videotape versions of the film. The first version showed only the fear provoking scenes included in the film. Narrated by Bonnie Franklin, star of the television program One Day at a 771me, this version showed scenes designed to scare viewers so they would stop smoking. These scenes included an interview with a young woman undergoing chemotherapy for lung cancer, sequences showing how smoke deteriorated the tissues of the skin, and a presentation by a doctor of the results of medical tests that demonstrated the effects of cigarette smoke on the unborn.

The second 15-minute videotape version included the most dramatic, fear-provoking scenes used in the first version, but also included about the 5 minutes of information on how to stop smoking. These scenes gave information on smokers' support groups and how the body recovered once a smoker quit.

The two video versions of the motion picture were evaluated several times during production. They were also evaluated by subjects during the experiment and in all cases were judged to be generally of high and comparable technical quality.

After treatments were administered, subjects completed the Smoking Attitude Scale (SAS; Baer, 1966). The SAS is a 21-itern measure using a five-response, Likert-type scale. It is reported to have a reliability estimate of .84.

Results of Study #4. Results are reported in Table 34-5. It was found that both experimental treatments were successful at significantly influencing subjects' attitudes toward smoking. Subjects assigned to one of the two versions of the videotaped adaptations of The Feminine Mistake had more negative attitudes toward smoking after viewing treatments than did control subjects. There was no statistical difference found between subjects categorized as being either field dependent or field independent, nor was there a significant interaction between field dependence and treatment.

34.6.3.1. Discussion of the Four Studies. Earlier, two specific questions were posed that served as guides for the design of the four studies included in this research program.

Question I. Is there a hierarchy of media types related to effectiveness in delivering persuasive messages? It seems obvious that media can be used to effectively deliver persuasive messages. Studies 1, 2, and 4 all reported attitude positions that were significantly different for subjects who viewed experimental treatments when compared to control subjects. There also seemed to be some evidence that motion pictures generally were more effective at persuading than were slide presentations. This conclusion was supported by the results of studies I and 2.

The impact of realistic persuasive messages on attitude change has been studied by social psychologists for over 4 decades. Reinforcement theory is based on the assumption that realistic messages have more cues for the viewer, and thus are more effective at persuading. Specifically, since motion pictures have more visual information, theoretically this should make motion pictures more effective than still pictures at persuading. The results from these studies seem to support the assumptions of this theory. While no hierarchy of media types could be developed based on this series of experiments, it does seem that this question should be investigated further.

Question 2. Does learner aptitude interact with media type when attitude change is the goal of a message? Based on the results of studies I and 2, there may be a relationship between field independence and filmed persuasive messages for the topics of soil conservation and hiring the disabled. Field-independent subjects seem more likely to be influenced favorably. It also seemed that films were, in general, better than slide presentations at delivering messages that changed attitudes.

34.6.3.2. Conclusions from the Four Studies. A fundamental assumption of the research presented above was that attitude change is an important concern of the educator, and that if attitudes are important, information on how attitudes might be formed or changed with media is needed. These four studies were conducted to examine the use of media to deliver persuasive messages. The results of the four studies tended to support the following conclusions: First, attitudes toward educationally relevant topics, such as conservation, smoking, and disabled persons, can be modified by using persuasive messages delivered by media. Next, it appeared that some types of media may be more effective than others at delivering information designed to change attitudes. Motion pictures seem to be more effective than slides. There also seems to be sufficient evidence to warrant further investigation into the relationship between the content of persuasive messages, the media used to deliver those messages, and the characteristics of the students who view the message.

Realistic media, such as films and video, were reported most often to be vehicles for delivering attitude change messages. The next step in Simonson's research agenda was to query filmmakers about the techniques they used when they designed persuasive motion pictures.

Study #5. Persuasive Films: Techniques Used to Change Attitudes (Simonson, 1985; Dimond & Simonson, 1988)

Purpose of the study. The four studies reported above indicated that persuasive films can change attitudes. The next study attempted to determine how films were planned and produced, and what makes it possible for a film to persuade a viewer to accept an attitudinal position.

Alfred Hitchcock is supposed to have once remarked to an executive producer that he never looked at motion pictures, to which the producer replied, "But where do you get all your ideas then?" (Rose, 1963). "How they did it" in a film starts a description heard in many film production conferences. The techniques used by one filmmaker in a successful film are often used as models for similar films. Filmmakers often produced films by formula. Many times they do not even realize what formula they are following (Rose, 1963). Most often, filmmakers decide on, or are hired, to present a position in a film. The filmmaker then works backwards, planning the presentation to include, through emphasis and selection, the ideas and techniques that would be most likely to elicit the desired reaction in the viewer. In other words, the desired attitude is identified, then filmmakers decide how to persuade the audience to accept this attitude as a consequence of viewing their film (Rose, 1963).

The identification of the specific techniques used by the filmmaker to accomplish these persuasive goals was the purpose of this study. Simonson's (1979) guidelines were used in the development of the questions asked.

Procedures. In order to obtain information from filmmakers about persuasive films, a Film-Makers Survey (FMS) was developed. A pilot version of the FMS was sent to a small sample of filmmakers and a revised version based on their comments and suggestions was developed. This revised FMS had two parts. Part I dealt with the filmmaker's background and experience. Part 2 asked the filmmaker to rate, discuss, or evaluate techniques used in persuasive filmmaking. Each item in part 2 of the FMS was directly related to one of the six design guidelines identified by Simonson (1979).

The catalog of the Council on International Non-Theatrical Events (CM) listed 150 filmmakers who were sent a copy of the FMS with a cover letter explaining the purpose ' of this study- These filmmakers were sampled because their films were listed in the CINE catalog as Golden Eagle Award winners and because the accompanying descriptions of their -films seemed to indicate that their motion pictures were persuasive. A random selection of cinematographers was not considered appropriate because the purpose of this study was to have individuals with experience in persuasive filmmaking evaluate techniques used to produce this type of film. Experience in persuasive filmmaking was necessary in order for a person to be able to satisfactorily complete the FMS. Included. with the questionnaire and cover letter was a stamped, addressed, return envelope. No follow-ups to filmmakers who failed to return the questionnaire were attempted, because the cover letter stated that if addressees had not produced persuasive films, they should disregard the FMS. Because subjects were selected based on a two- or three-sentence description of one of their films in the CINE catalog, it was very likely that many of them were not actually persuasive filmmakers and, consequently, discarded the FMS.

Results. Of the questionnaires, 51 were returned, for a response rate of 34%. This fairly small percentage of returns was expected because of the technique used to select subjects and because no follow-ups were attempted. A careful analysis of the completed questionnaires did not reveal any discernible, confounding pattern. An analysis of the known characteristics of those filmmakers who completed the survey as compared to those who did not return it failed to reveal any significant relationships that might have indicated that a biased subset of filmmakers answered the FMS as compared to those who did not.

Background information on the responding filmmakers. All 51 of the filmmakers who responded indicated that producing motion pictures was their primary method of employment. The average length of time respondents had been employed as filmmakers was slightly more than 18 years. The range of years of employment was I to 40. Of the respondents, 91% were males. Their average age was 46. Ages ranged from 25 to 68 years.

The average number of films of all types produced by each filmmaker was 142. The average number of persuasive films produced was 29 (range = I to 200). For persuasive films, the average film length was 18.3 minutes. The shortest persuasive film was reported to be 2 minutes. The longest was 30 minutes. There were a large number of 28minute films reported, probably because this was a popular length for films that were to be broadcast. Filmmakers reported working in production companies with an average of 11 employees. The smallest company was a one-person freelance operation. The largest company employed 35. .

Of the filmmakers, 37% reported that they had no formal school training in filmmaking, only on-the-job training; 16% had some college experience. Those who had masters degrees or more were 21%; 5% reported that they had attended a trade school. Only 9% reported having any formal training in producing persuasive films.

The definition of persuasive film used for this study was considered appropriate by 78% of the filmmakers who answered the FMS. Almost without exception, those who did not like the definition thought that it was too narrow and should be expanded to include broadcast, noneducational uses of films. The definition used for this study was: "A persuasive film is a training or educational film that has influencing, persuading, or changing of attitudes as its primary purpose-

Most respondents thought that the market for persuasive films would increase in the future (73%), and that about 40% of the educational film market was for films that primarily persuade rather than inform. The average cost of a 10-minute persuasive film was estimated at $29,000, or $2,900 per minute (range = $13,000 to $65,000). This price was estimated as being only slightly higher than the cost of an informative film of the same length.

Persuasive-film production techniques. One of the major goals of this study was to determine how filmmakers would go about producing a film when persuasion was their goal Filmmakers responding to the FMS considered persuasive films to be planned and produced a little differently than other educational films (X = 3.79; 5 = very differently; I = exactly the same). One major difference was the importance of a prescript writing target audience assessment that most filmmakers considered critical to the success of their persuasive films (X = 4.27; 5 = critical; I = not necessary). Technical quality of persuasive films was considered important, but only slightly more so than for any film (X = 3.59; 5 = critical; I = less important than for other films). The "outs," the percentage of film not used, for persuasive m6tion pictures was estimated at being only slightly greater than for informative films (X = 3.59; 5 = much greater, 3 about the same; I = much less).

To determine the production techniques considered most effective for persuasive films, several somewhat overlapping groups of procedures were presented in the FMS for the filmmakers to rate. An analysis of this rating process follows.

When asked to rank three general statements concerning how important each was to persuasive filmmaking, the filmmakers considered the need to arouse the viewer emotionally or to promote some reaction in the viewer relative to the content of the persuasive film as the most important of the three presented (X = 1.23; 1 = most important; 3 = least important). The technical quality of the film (X = 2.37) and the need to present considerable information in the film about the topic (X = 2.31) were not considered as important as involving the viewer in the message of the motion picture.

Because the production of persuasive films is considered controversial by some, several questions were included in the FMS to determine what filmmakers thought about the propriety of producing films that were meant to persuade rather than to inform. Most considered persuasive films to be much more exciting to produce than other types of motion pictures (X = 4.25; 5 = much more exciting; I = not much fun). Only 10% reported having problems with the morality of producing attitude change films. However, most indicated that they would refuse to produce a film that was intended to promote a position they did not believe in.

Based on the FMS results reported above, it would seem that there were several ingredients agreed on by the filmmakers as being likely to promote attitude changes when they were included in the planning or production of persuasive films. Techniques considered important for successful persuasive films were that they should:

• Be planned and based on the results of a target audience assessment
• Arouse and promote in the viewer some action, either intellectual or behavioral
• Present new information to the viewer on the topic Be realistic and believable to the viewer
• Include motion and action in scenes rather than static visuals

• Be enjoyable, or fun, to watch Be as nonverbal as possible
• Be used with follow-up discussions as outlined in a teacher's guide that should accompany the film

This hypothetical persuasive film should be no longer than necessary and have a budget of about $3,000 per minute. Technical quality would be important for this motion picture, but only slightly more so than for any film. The "outs" ratio would probably be slightly greater for this film than for a regular informative motion picture of the same length.

Techniques not considered important or effective for persuasive films were the use of considerations of-

• Color rather than black and white

• Film length

• Graphs, charts, and other written scenes

• "Scare" tactics that would attempt to show the dire consequences of not adhering to the message of the persuasive film

• Talking faces

• Animation

• Inspirational messages

Generally, it was thought that an effective persuasive film was one that was believable and realistic, presented new information, was fun to watch, promoted involvement or action in the viewer, was more visual than verbal, and was used correctly by the teacher. These results were the amalgamation of those found in two studies (Simonson, 1985; Dimond & Simonson, 1988).

Study #6. Subliminal Messages, Persuasion, and Behavior Change (Treimer & Simonson, 1988). In 1988, Treimer and Simonson published the results of a slightly different type of study that attempted to investigate the impact of subliminal messages on the level of emotional involvement felt by the viewer of videotapes. It was felt that subliminal messages, if effective, might provide an alternative method of intellectually or emotionally involving the viewer of a videotape in the persuasive message, as proposed by guideline #6 (Simonson, 1979).

Subliminal perception was defined as any word, image, or sound that is not perceived within the normal range of consciousness but that makes an impression on the mind. This phenomenon involves words or pictures that are flashed so quickly that the eye cannot transmit them to the conscious brain, or words spoken at such a volume that they evoke no conscious memory.

The purpose of the Treimer and Simonson (1988) study was to determine whether viewing a commercial videotape containing written and aural subliminal messages was more effective at producing weight loss and attitude change toward weight loss than a videotape containing the same visible message but with no subliminals. Weight Loss Video Programming, by Hypnovision, Inc., was the videotape selected for this study. It endorsed no specific weight loss or exercise plan and required nothing of viewers other than their willingness to change diet and exercise habits and to watch the videotape that contained subliminal messages daily for 30 days.

Participants were measured to see if changes occurred in the following areas:

on the position that "the likelihood that a receiver will accept the conclusion advocated in a given lesson is in part a function of the receiver's perception of the source's or model's credibility" (p. 286).

I . Food and Exercise Attitude (FEAT; r = .72). This test of attitudes was administered at the beginning and end of the treatment period.

2. Food Intake Recall (FIR). FIR was measured by using a 1-day recall of food intake at the end of the testing period.

3. Weight and Skinfold Test (WST). WST was measured at the beginning and end of the treatment period.

It was hypothesized that participants who viewed the videotapes with subliminals would change their attitudes toward dieting, would have a healthier food intake, and would lose weight. A pretest/posttest control group design was used. Participants were volunteers who were randomly assigned to one of two treatment groups: video with subliminals or video with no subliminals. Both groups watched

the 22-minute video at least 25 times in a 35-day period. Two versions of the video were prepared, one with the subliminals and one without.

Results of Study #6. The results of the study showed that FEAT scores improved for the group watching the video with the subliminals (11.8 vs. 7.37), but that the difference was not statistically significant. Results of the FIR and WST testing showed little numerical difference in the scores of participants in either group. It was concluded that subliminal messages did not appear to have an impact on either attitude or behavior.

This study is included in this review because it represents a slightly different approach than the five studies summarized previously. The results of this study effectively closed the door on research on subliminals. As a matter of fact, several months later Hypnovision, Inc., stopped marketing subliminal videos.

Fleming and Levie proposed a series of attitude change principles in 1978 that were based on their comprehensive review of the media and attitude research. Their discussion had three categories of information. The first was based on the classic SMCR model of communication (see 4.3). (The Source presents a Message through a Channel to a Receiver). They also developed principles related to modeling of appropriate behaviors, and they concluded with principles related to creating and managing dissonance. In 1993, Fleming and Levie included an updated and revised series of attitude change principles (Bednar & Levie, 1993) that are included next. SMCR Principles-the Source. The first series of guidelines proposed by Bednar and Levie were concerned with the source of the persuasive message and were based

Principle I. L High-credibility sources exert more persuasive influence than low credibility sources.

Principle 1.2. Sources perceived by the receiver as attractive are more influential.

Principle 1.3. The quality and structure of the arguments in a persuasive message are more critical for credible sources than for attractive sources.

Message Principles. Next, Bednar and Levie identified principles that were concerned with the content of the persuasive message. Message principles were considered to be closely related to source-based principles.

Principle 1.6. Arguments are more effective if they are relevant to the receiver’s needs.

Principle 1.8. It is almost always advisable to state the conclusion explicitly rather than to allow receivers to draw their own conclusions.

Principle 1.9. Repetition helps, but only one or two repetions are likely to have any additional effect.

Principle 1.4. Be sure the receiver is informed of the expertise of a high-credibility communicator.

Principle 1.5. To enhance communicator attractiveness, establish belief congruence with the receiver by arguing in favor of positions die receiver is known to hold.

Arguments are more effective if they are relevant to the receiver's needs.

Principle 1.7. Generally, two-sided arguments are slightly more effective than one-sided messages.

It is almost always advisable to state. the conclusion explicitly rather than to allow receivers to draw their own conclusions. Repetition helps, but only one or two repetitions are likely to have any additional effect.

Channel Principle. A principle related to the channel of communication was presented next by Bednar and Levie. Channels were explained to refer to both media and to senses.

Principle 1.10. No one media type has been explicitly shown to have greater persuasive effectiveness than any other media type. Face-to-face communication, however, is more effective in promoting acceptance than mediated communication, particularly in difficult cases.

Receiver Principle. Finally, Bednar and Levie proposed one principle related to the receiver of a persuasive message. They stated that is very important for the designer of the attitude change lesson to know as much as possible about the student who will receive the lesson.

Principle 1. 11. It is very difficult to change the attitudes of receivers who are highly committed to their positions on an issue.

Modeling Principles. Next, Bednar and Levie proposed five principles that dealt with the use of modeling as an instructional strategy and with the credibility of the model.

Principle 2. 1. High-credibility models exert more persuasive influence than low-credibility models.

Principle 2.2. In order for modeling to be effective, the learners must comprehend the presentation as a demonstration of specific behaviors.

Principle 2.3. In addition to observing the model demonstrating the behavior, learners should observe the model being reinforced for that behavior.

Principle 2.4. Role playing can have a powerful

persuasive impact.

Principle 2.5. Active participation produces more attitude change than passive reception of information.

Dissonance Principles. Last, Bednar and Levie offered six principles for creating and managing dissonance in order to produce attitude changes. Their principles were based on Festinger's (1957) cognitive dissonance theory.

Principle 3. 1. If a person can be induced to perform an important act that is counter to the person's own private attitude, attitude change may result.

Principle 3.2. When a person is induced to perform an attitudinally discrepant act because of promise, of reward or punishment attitude change will occur only to the extent that the person feels that the magnitude of the reward or punishment was insufficient to justify the attitudinally discrepant behavior.

Principle 3.3. A message should demonstrate the social acceptability of the desired attitude and the reward available socially for behavior consistent with the attitude.

Principle 3.4. The message should alternate, between presenting information discrepant with existing beliefs and inducing behaviors discrepant with existing attitudes to maximize dissonance.

Principle 3.5. Attitude change lessons should be

structured so that attention is paid to

the cognitive (information), affective

(feeling I ), and behavioral (acting) elements of the attitude.

Principle 3.6. Messages should use approximations to move attitudes gradually between a current status and a desired state.

Bednar and Levie (1993) included with each principle a discussion of the literature that supported it. The Bednar and Levie principles are practical and effective, and provide

considerable guidance to the designer of persuasive messages. They stated in their conclusion that "common to all of these (principles) are opportunities for free choice and control by students, opportunities for success, and lessons which present and confront alternative perspectives" (p. 302).

Simonson's six studies, Bednar and Levie's 22 principles, and other research reported during the 1980s and 1990s have provided considerable information about attitude change and instructional technology. The next section of this chapter will propose a set of guidelines for designing mediated messages that persuade. These guidelines are based on the research reviewed above and on the other studies that were found in the literature. It is important to note that revisiting this research area produced no startling changes from the reviews reported a decade and a half ago. Attitude change is quite predictable if mediated instruction is correctly designed. Media, at best, play a minor role in persuasion when compared to the message delivered by the medium or the methodology of instruction.