v a». ' .jlafl.\}|‘\.k r ‘\"'~Z';wvfl” r} mm: 25¢ per day per ital RETURNING LIBRARY MATERIALS: P‘lace in book return to runove charge from circulation records AN EXPERIMENTAL STUDY OF THE EFFECTIVENESS OF DOCUMENTARY FILM AS AN ATTITUDE INFLUENCE DEVICE ON ELEMENTARY SCHOOL PUPILS By Peter J. Brobeil A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Elementary and Special Education 1981 C) Copyright by PETER J. BROBEIL 1981 ii ABSTRACT AN EXPERIMENTAL STUDY OF THE EFFECTIVENESS OF DOCUMENTARY FILM AS AN ATTITUDE INFLUENCE DEVICE ON ELEMENTARY SCHOOL PUPILS By Peter J. Brobeil The purpose of this study was to evaluate the effectiveness of documentary film as an attitude influence device on elementary school pupils. A model for persuasive message design was developed and applied to the production of a 16mm color/sound documentary film entitled Circle Song. The film portrays the eastern grey wolf's ecological niche within its natural habitat. Narrated in a Native American language, the film utilizes English in its translation. A pretest-posttest control group design was utilized within the study. Participants were selected from grades '3, 4, and 5, as suggested by the findings of the Messerschmidt study, first reported in 1933. Prior to the experimental phase of the study, a pilot survey was con- ducted with a similar age sample. Results of the survey guided the design and production of the 16mm experimental film. Results of a pretest were paired, and random assignment to a treatment or control group was conducted for each grade level participating in the study. The experimental proce- dure involved the projection of the test film and subsequent Peter J. Brobeil testing for resultant affective disposition and cognitive learning. The test instrument consisted of a forty-item Likert Scale, developed with the assistance of a panel of judges. Test items were randomly redistributed for each testing session. Testing of seven hypotheses utilized a linear regression model. All hypotheses were tested using t statistics with a confidence level established at c = .05. Results of this investigation indicate that significant (a = .05) cognitive change had occurred within each experi- mental group tested. Significant (c = .05) affective change was recorded at the third and fifth-grade experimentallevel. No significant (c = .05) affective change was noted at the fourth-grade level. It is suggested that the recorded dis- parity in gain score concerning affective elements might be attributed to developmental factors. No "sleeper effect" or "decay" was recorded on a delayed posttest administered ten days after the experimental treatment. Dedicated To My Parents John Thomas and Phyllis Friant BROBEIL ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to all of those persons who contributed toward the completion of this investigation. Special acknowledgement and thanks are given to Dr. Bruce Cheney, Chairperson of the Doctoral Committee and Director of the Dissertation. His positive involvement was instrumental in guiding the project to a successful completion. To Dr. M. Ali Issari, who made me a filmmaker and who greatly assisted the researcher throughout all aspects of the experimental film production, I am indebted. Special thanks are extended to committee members Dr. James Page, Dr. Eckhart Dersch, and Dr. William Walsh for critiquing the document and for offering constructive sug- gestions concerning the project. To George Cornell, I owe a debt of friendship. He willingly involved himself in the project, serving in the role of Native American consultant. Mrs. Cynthia Pitsinger of St. Louis, Missouri, and Dr. Harry Frank of the University of Michigan graciously allowed their wolves to be filmed for the project. To them, a special thank you. iv V Thanks to Thom Alcoze of the Bidasgiking Settlement, St. Charles, Ontario, Canada, for portraying Anishnabe within the film. To George Roy, for his Ottawa narration, . also special thanks. Bea Peters arranged for the Native music score used in the film. It was performed by The Night Nolf Singers of Lansing, Michigan. To these gracious people, I also extend my thanks. The original guitar duet, created as the main music track for the film, was performed by Peter William. Pat Cyccone volunteered his expertise and time, serving as audio consultant for the film. To these gentlemen, a sin- cere thank you is extended. To Lola D'Meara, principal of the Pinecrest Elementary School in East Lansing, Michigan, and to her staff and pupils, a special note of thanks for facilitating the experimental procedures. Also, an expression of gratitude is in order to these kind people for extending a high level of hospitality and cooperation toward the researcher. You made me feel at home. Michael Cook, a fellow devotee of things natural and wild, took many photos used within the experimental film. His statistical and mathematical considerations also greatly influenced this research. A final acknowledgment and sincere thanks is extended to Craig Goettsche and Donald Nyhowski. Their invaluable assistance during the analysis made the dissertation vi possible. To them, the Cyber 750 is but a servant. To this mortal, friends like Michael, Craig, and Don are vital. TABLE OF CONTENTS LIST OF TABLES ........................................ xi CHAPTER I. THE PROBLEM ............................... 1 Identification of Problems ....................... Societal Incorporation ........................... 2 3 Purpose .......................................... 4 5 7 Importance ....................................... Generalizability ................................. Initial Considerations ........................... 8 Statement of Research Questions and Hypotheses ... 12 Research Questions ............................... 13 Research Hypotheses .............................. 13 Assumptions ...................................... 14 Limitations ...................................... 17 'Internal Validity ............................. 17 Sources of External ValidityANhich May Limit the Study ..................................... 17 Definition of Terms .............................. 18 giganization of the Thesis ....................... 20 CHAPTER II. REVIEW OF RELATED LITERATURE ............. 22 Introduction: Major Areas of Review .............. 22 Source ........................................... 23 Credibility ................................... 23 Attractiveness ................................ 29 Message .......................................... 34 Hovland, et al., on Fear ...................... 34 Fear Threat Factors ........................... 41 Audience ......................................... 48 Medium ........................................... 59 Effect ........................................... 64 Summary .......................................... 67 Source Summary ................................ 68 Message Summary ............................... 7O Audience Summary .............................. 74 Medium Summary ................................ 76 Effect Summary ................................ 77 CHAPTER III: DESIGN OF THE STUDY ...................... 79 Description and Selection of the Sample .......... 79 Pretest .......................................... 81 Independent Measure .............................. 82 Experimental Film ............................. 82 Dependent Measure ................................ 84 Test Instrument ............................... 84 Experimental Design .............................. 86 Design Over Variables ............................ 88 Procedures for the Investigation ................. 88 Preliminary Procedures ........................... 88 Collection of the Data for the Study ............. 93 Posttest ...................................... 93 Delayed Posttest .............................. 96 ix Data Estimation Procedures ....................... 98 Sorting Prpgram ............................... 98 Time Series Processor (TSP) ................... 99 Hypotheses Testing ............................ 101 Hypotheses and Data Analyses ..................... 102 Summary .......................................... 103 CHAPTER IV: ANALYSIS OF DATA AND FINDINGS ............. 105 The Sample ....................................... . 105 Hypotheses, ....................................... 106 Results of Hypotheses Testing .................... 107 Hypotheses 1 and 2 ............................ 108 Hypotheses 3 and 4 ............................ 113 Hypotheses 5 and 6 ............................ .118 Hypothesis 7 .................................. 120 Conclusions for Hypotheses 1 and 2 ............ 122 Conclusions for Hypotheses 3 and 4 ............ 126 Conclusions for Hypotheses 5 and 6 ............ 126 Conclusions for Hypothesis 7 .................. 126 Summary .......................................... 126 CHAPTER V: SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ... 130 Conclusions ...................................... 137 Discussion ....................................... 140 Implications ..................................... 142 Recommendations for Further Investigation ........ 144 BIBLIOGRAPHY .......................................... 146 APPENDICES ...... Appendix Appendix Appendix Appendix Appendix Appendix Appendix H. Appendix I. A B C Appendix D. E F G ...................................... 163 PILOT STUDY RESULTS ................. 163 PILOT STUDY OPINION SURVEY .......... 171 ANSWERS TO PILOT SURVEY .............. 176 PRETEST - TEST #1 ................... 178 POSTTEST - TEST #2 .................. 181 DELAYED POSTTEST - TEST #3 .......... 185 LETTER TO PARENTS OF CHILDREN PARTICIPATING IN STUDY .............. 189 SCENES AND NARRATION OF CIRCLE SONG 19D MODEL: PERSUASIVE MESSAGE DESIGN .... 199 LIST OF TABLES- 3.1. Population of Pretest ............................ 82 3.2. Pretest/Posttest Control Group Design ............ 86 3.3. Variable Matrix .................................. 88 3.4. Population for Posttest .......................... 94 4.1. Data for Ho: 1 and 2, Grade 5 .................... 110 4.2. Data for Ho: 1 and 2, Grade 4 .................... 112 4.3. Data for Ho: 1 and 2, Grade 3 .................... 114 4.4. Data for Ho: 3 and 4 ............................. 117 4.5. Data for Ho: 5 and 6 ............................. 121 4.6. Data for H3: Grade 5 ............................. 123 4.7. Data for H3: Grade 4 ............. ................ 124 4.8. Data for H3: Grade 3 ............................. 125 4.9. Descriptive Information .......................... 128 4.10. Summary of the Findings .......................... 129 xi CHAPTER I THE PROBLEM Since the publication of Rachel Carson's Silent Spring in 1962, man's relationship to his natural environment has been receiving increased national and international atten- tion. The United States' past laissez-faire economic orien- tation concerning natural resources is beginning to be questioned. Aldo Leopold1 cautioned in 1948 that: There is yet no ethic dealing with man's relationships to the land and to the animals and plants which grow upon it. Land . . . is still property. The land relationship is still strictly ecogomic, entailing privileges, but not yobligations. Bryan (1974) concluded that the failure of environmental change methodology can be blamed on the predominate use of "rational education strategy models." Based upon concepts of "truth-love" and the "perfectibility of man," these models assume that man is rational: that he not only chooses what is best for himself, but for society as well. Bryan 1Aldo Leopold: 1887-1948. Co-founder of the Wilderness Society; founder of first Forest Wilderness Area in the United States, 1924; founder of professional game management concept. Forester, naturalist, advisor on conservation to United Nations, noted professor and author. 2Aldo Leopold, A Sand County Almanac (New York: Ballantine Books, 19667. 1 2 argues that due to man's history of seeking short-term material self-interest, he demonstrates an irrational nature. This concept of man continues to influence his decision-making processes.3 It is therefore necessary for any individual under- taking environmental change to: . . . fully understand the nature of social power and the means by which it is acquired. Unfortu- nately there is a dearth of social power models of change that environmental activists can use to create social power bases that bring about desir- able changes, primarily because such models incorporate "distasteful" issues like conflict, manipulation and influence. Yet what major environmental victories have not been characterized by those dimensions of soc1al power? Identification of the Problem That man has often had a less than mutually beneficial relationship with the environment is a fact with which few persons would disagree. It could be argued additionally that as a species we now stand at a crossroad in time.- Perhaps, at no other period has man had such an absolute capacity to destroy the delicate strands that weave the ultimate support system into one universal whole. Based upon present trends and a less than immaculate past, it would further appear that our educational processes 3William L. Bryan, "Toward a Viable Environmental Movement," Journal of Applied Behavioral Science, vol. 10, no. 3 (1974). p. 392. 4Ibid., p. 393. 3 must do much more than prepare a society of academicians. Attributes of awareness, critical thought and analysis, understanding, and humane application must be paramount in our search for solutions that could well determine the future for our societal existence. Societal Incorporation The normal and healthy human infant, once physically and emotionally ready, can begin to be assimilated into its society through the application of various message stimuli. As the stimulus situation continues over time and with proper intensity, the receiver begins to think, feel, and respond in the desired manner. The child begins to acquire feelings, values, beliefs, and attitudes about his/ her world, thus assuring the perpetuation of the societal group of which he/she is a part. Basic attitude components are complicated parts of each individual and often exist as unrecognized entities by their holders. Allport (1935) defines attitudes as, "a mental or neural state of readiness to respond, organized through experience and exerting a directive and/or dynamic "5 Therefore. the way we behave may influence on behavior. well indicate our underlying attitudes. Demonstrated societal behavior appears to indicate that many held attitudes are in conflict with co-existing life 5Gordon W. Allport, "Attitudes," in A Handbook of Social Psychology (Worchester, Mass: Clark University PFess, 1935), p. 810. 4 forms. In addition, these conflicts may threaten our own continued existence and must be resolved while time remains. Purpose From an accumulation of research evidence, as presented in Chapter II, it would appear that the present adult generation will not be able to solve these problems. Therefore, we need to turn to future generations and develop better strategies. Ultimately, these new strategies must alert the youth to the crisis, develop ecologically positive attitudes within the young, and persuade them to become behaviorally involved in the solution search. This study is concerned with one aspect of the search for better strategies of instruction. Because attitudes play such a vital role in the total educational process, an attempt will be made to develop a procedure for affecting, instilling, or shifting attitudes of elementary students, such that the post-treatment attitude will be more compati- ble with the procedural goal of the researcher. Prior to this study, little appears to have been accomplished in the area of documentary film as a device for affecting attitudes of elementary school children. It is generally recognized that this medium is highly successful in the transmission of content knowledge, but as an attitude influence device, little appears settled as to its effec- tiveness. We do know, however, that cinema has been demon- strated to facilitate factual learning at an accelerated 5 rate, twenty-two percent greater than some types of classroom instruction.6 From an abundance of research data concerning the processes of attitude influence, it appears plausible to apply specific principles of attitude influence strategy to the production of a 16mm documentary film, such that the resultant product might contribute additively to the contin- uing investigation of the attitude influence phenomenon. An attempt will be made to evaluate the effectiveness of the medium upon viewer attitudes concerning subject content and the attitude object. In addition, the study will investi- gate several of the properties, attributes, and limitations of the medium in an effort to maximize audience identifica- tion and acceptance of its message. The procedural steps involved in the investigation may demonstrate a model of social influence. Prior to this investigation, it appeared that educational film productions rarely utilized such models. Importance Part of the educator's responsibility to society is to prepare our students to assume ultimate stewardship of all systems that enable the human species to endure. This should be accomplished in the most beneficial and least detrimental manner to the system as a whole. 6Roy Paul Madsen, The Impact of Film (New York: MacMillan Publishing Co., 1973), pp. 485-498. 6 It would appear that various species, including man, have received substantial abuse, partially as a result of erroneous information presented by high credibility sources, culminating in change-resistant attitudes. One such species, the grey wolf (canis lupus), seems even today to be particularly subject to negative attitude formation. Chil- dren are exposed at a very early age to myths, tales, and legends which contribute to an attitude set that conflicts with the survival outlook for the species. Generally, the myths, tales, and legends are unfounded in fact and could result in, or contribute to, the development of a permanent negative attitude set spanning one's life. Canis lupus, therefore, offers the experimenter an opportunity to investigate an area of attitude formation with a population still acquiring its initial attitude set. Maximum suggestibility concerning attitude change has been identified as occurring between eight or nine years of age.7 1 Research from the areas of social psychology, advertis- ing, marketing, communications, and media suggests that specific strategies could be utilized in the production of cinemagraphic media which would increase the ultimate impact of content learning and attitude influence. It therefore 7Theodore X. Barber and David S. Calverly, "Hypnotiza- bility, Suggestibility, and Personality: A Study with the Leary Interpersonal Checklist," British Journal of Social and Clinical Psycholo 3 (1964): 149-150. Ramona Messer- schmidt, "The Suggesti ility of Boys and Girls Between the Ages of Six and Sixteen Years," The Peda o ical Seminary and _Jgurnal of Genetic Psychology 3 (1933): 323-437. 7 appears plausible to produce a film that would not only maximize content learning, but would have a greater impact on the target's attitude set. This impact might very well shift the questioned attitudes more in line with the film maker's ultimate goal, when strategies concerning source, message design, and audience characteristics are judiciously applied as an integral part of the design and production phase of the film. The final reason for undertaking this investigation is to develop a model which, when applied to any of numerous social issues, might provide a greater probability ofachiev- ing the desired result, whether it be cognitive or affective in nature. Generalizability . The findings of this study are expected to contribute to a growing understanding of the attitude phenomenon throUgh empirical research. In addition, as the procedures for attitude influence will be incorporated within the construct of a 16mm documen- tary film, it is expected that a model for future educa- tional film productions will emerge. The model for the test film production in addition may have applicability to other procedures and strategies for social change, general educa- tion, and/or media production for educational purposes. 8 Initial Considerations (1) On February 12, 1980, a pilot study was conducted at the Pinecrest Elementary School in East Lansing, Michi- gan. The purpose of this study was to aid in a determina- tion of factual knowledge and existing attitudes held by the test population toward the subject species, the North Ameri- can wolf. The sample consisted of twenty-five third grade, twenty-five fourth grade, and sixty-eight fifth grade children. The survey was structured as a forty-three-item test utilizing a Likert scale, with each question having three possible responses. The answers indicated the students' agreement, neutrality, or disagreement with each test item. A panel of judges, exhibiting expertise in the topic areas, predetermined a correct response for each question and verified each measure. Student responses were compared with those of the judges. The survey was evaluated on a grade level basis, with each respondent remaining anonymous. Grade raw scores were converted to percentages of class answering for each possible item alternative. The survey included nine categories, which indicated a level of knowledge and attitude with respect to wolves, ani- mals in general, authority figures, and environment. The nine categories consisted of: cooperation in answering indi- cators, knowledge of animals, attitude toward animals, knowledge of wolves, attitude toward wolves, goodness/bad- ness of animals, attitude toward environment, attitude 9 toward message source/authority figures, and degree of exposure to wolves. Results of the pilot survey were used to guide the selection of knowledge and attitude components for inclusion in the design of the 16mm experimental film. Additional findings and the pilot survey are included in Appendices A-C. (2) The 16mm experimental film, used as the main treatment procedure within the study, was designed in accordance with results obtained from the pilot survey. During development of the narrative script, care was taken to assure a level of comprehension compatible with the youngest children participating in the experiment. The Fry Test of Readibility was applied to the narrative script, and it was determined that the level of reading complexity approximated that of a second-grade reading level.8 Assuming that the average attention span of a third- grade child approximates fifteen minutes in length, while the fifth-grade child approxiamtes a thirty-minute attention span, the experimental film length was limited to twenty- four minutes. With a high interest content level included within the film, it was estimated that the majority of par- ticipants at each grade level would view the film while operating within their level of attention. 8Graph for Estimating Readability by Edward Fry, Rutgers University Reading Center, New Jersey. From Language Arts (September 1975). 10 The content of the film was developed in direct accordance with the results of the pilot tést. Information that was well known to the subjects was not included within the film. Topics relevant to the investigation, that reflected confusion or misunderstanding when presented within the pilot survey, were reintroduced within the experimental film. (3) Prior to the experimental aspect of the study, it was necessary to apply to the University Committee on Research Involving Human Subjects (UCRIHS) at Michigan State 9 0n University for permission to conduct the investigation. September 8, 1980, the UCRIHS approved the pending investi- gation citing that "the rights and welfare of the human sub- jects appear to be adequately protected...." Upon approval of the UCRIHS, the experimenter then contacted the Assistant Superintendent for Instruction and Professional Development for the East Lansing, Michigan, Public Schools. Permission was sought and granted to offi- cially conduct the research within the East Lansing School District. As the Pinecrest School was the specific institution in which the experiment was to be conducted, procedures were discussed with the principal, and permission to utilize the facility and students was granted. 9The university committee oversees and assures that planned investigations are in compliance with the best interests of the subjects and are non-detrimental in nature. 11 (4) The research was conducted with third, fourth, and fifth grade students. Research in attitude processes appears to fix the chronological age for maximum attitude ‘0 Utilizing the three suggestibility at or about age nine. grades enabled the experimenter to bracket the specific test grade to allow for any differences that might exist in the population and to afford a means of comparison for the findings. (5) On May 18, 1981, a meeting was held with the administration and participating faculty at Pinecrest Elementary School. During this meeting, the participating teachers were briefed on the experimental design, and proce- dures for conducting the research were established. (6) One week prior to the investigation, a form letter explaining the experiment was sent home to the parents. All selected students were granted permission by their parents to participate in the investigation. A copy of the form letter is included as Appendix G. (7) Each participating grade level was administered a pretest to ascertain a cognitive and affective score rele- 11 Scores were then vant to the experimental subject. paired, and random assignment to a control or treatment group was conducted. 1oMesserschmidt, "Suggestibility of Boys and Girls." 11See Appendix D. 12 Statement of Research Questions and Hypotheses The main impetus of this research was to determine the ultimate value of documentary film as an attitude change device when designed and produced in accordance with spe- cific attitude influence strategies. The investigation included a test population composed of third, fourth, and fifth grade elementary school pupils. As indicated previ- ously in this study, research indicates that this age group is most susceptible to attitudinal influence. If the composite attitude set which one possesses in life is primarily established during the early formative years, attempts at creating positive ecological attitudes within that individual should be considered a valuable undertaking. Three general questions are posed in conjunction with this investigation. First, will the utilization of specific attitude influence strategies applied to the production of documentary film result in a significant attitude shift in a predetermined direction? Secondly, will fourth grade ele- mentary school pupils exhibit an attitude shift in the desired direction, greater than that demonstrated by either the third grade or fifth grade sample? Research evidence indicates that children in the eight to nine-year old cate- gory are most susceptible to attitude influence attempts. The third research question is quite similar to the first, but differs only in that it seeks to ascertain if there is a change in content knowledge. 13 Research Questions The following three questions compose the main impetus of this research undertaking. (1) (2) (3) Will the utilization of specific attitude influence strategies taken from the areas of attitude research, when applied to the production of a 16mm documentary film, result in a significant attitude shift in a pre- determined direction? Will fourth grade elementary school pupils exhibit an attitude shift greater than that of third or fifth grade elementary school pupils? Will the utilization of specific attitude influence strategies taken from the areas of attitude research, when applied to the production of a 16mm documentary film, result in substantial content learning when viewed by elementary school pupils? Research Hypotheses Answers to the above questions will be sought by testing of the following hypotheses: (1) (2) (3) Fourth grade mean gain between the treatment and con- trol group will be significantly greater than third and fifth grade mean gain concerning attitude change. Fourth grade mean gain between the treatment and con- trol group will be significantly greater than third and fifth grade mean gain concerning cognitive learning. There will be a significant difference between treat- ment and control groups, with respect to attitude 14 change, at each grade level tested. (4) There will be a significant difference between treat- ment and control groups, with respect to cognitive change, at each grade level tested. (5) Fifth-grade mean posttest treatment scores, with remxmt to attitude, will significantly exceed fourth-gradenman posttest treatment scores, which will significantly exceed third-grade mean posttest treatment scores. (6) Fifth-grade mean posttest treatment scores, with respect to cognitive learning, will significantlyexceedfourth- grade mean posttest treatment scores, which will sig- nificantly exceed third-grade mean posttest treatment scores. Each hypothesis was tested at the c = .05 level of statistical confidence. Assumptions The basic assumption of this investigation is that attL- tudes can be accurately assessed. Although researchewidence suggests that attitudes can be measured, the nature and com- plexity of attitudes pose difficulties for the researcher. It is generally accepted that attitudes are composed of three basic elements: affective, cognitive, and conative. Affective elements are believed to be emotional compon- ents of our make-up such as feelings, beliefs, etc. It is recognized that these elements may be evaluated by the uti- lization of several different measuring devices. The Likert Scale was the device utilized within this investigation. 15 One's accumulative knowledge with respect to a specific attitude object comprises the second attitude element. This knowledge element is perhaps the easiest to evaluate utiliz- ing any of numerous and popular methods. Within this inves- tigation, the Likert Scale was also utilized to measure cognitive elements.12 As the investigation was primarily concerned with evaluating the effectiveness of documentary film, the third element, the conative or behavioral component, could not be controlled. A second assumption concerns the ability of the ana- lytic procedures to make accurate measurements when the analysis was performed only upon two of the attitude ele- ments. This missing third element perhaps is the most accu- rate and valuable attitude predictor. It is generally accepted that behavior often is the manifestation of an underlying attitude or set of attitudes. Assuming that attitude components can be distinguished and independently measured, a Likert Scale was utilized to gather the experi- mental data concerning the affective and cognitive attitude elements.13 12Philip Zimbardo and Ebbe B. Ebbesen, Influencing Attitudes and Changing Behavior (Massachusetts: Aadison- Wesley Publishing Company, 1969). 13Paul B. Hournshell and Larry Liggett, "Assessing the Effectiveness of Environmental Education," The Journal of Environmental Education 5 (Winter 1973): 27-28. 16 A total of twenty-four statistical tests were performed upon the data. Hypothesis testing was conducted utilizing a regression equation, which was completely specified or defined by qualitative variables. A third assumption involves the acceptance of the test instrument (Likert Scale) as an accurate measure of cogni- tive and affective change occurring as a result of the experimental treatment. After careful evaluation of the pilot study data, cognitive and affective environmental components were organ- ized. These attitude components were later incorporated into a "story form" involving the attitude object. Subse- quently, the story form evolved into the experimental film. Utilizing the attitude and knowledge components incorporated within the experimental film, a panel of judges, exhibiting 'subject expertise, assisted in the development and verifica- tion of all test instruments. The fourth assumption concerns the degree to which the. experimental treatment (film) positively or negatively affected the experimental population. It was assumed that the various attitude and persuasive elements incorporated within the film were utilized properly in structuring the content, form, and intensity of the persuasive attempts. 17 Limitations The limitations of this study involve the degree to which the aforementioned assumptions are in fact valid. The investigation is limited in that factors involving internal. and external validity may not have been totally controlled. However, the experimental design incorporated within this investigation controls for specific sources of:14 Internal Validity: (1) History: Events occurring between the first and second measurement beyond the experiment. (2) Maturation: Processes within the respondent Operating as a function of time. (3) Testing: The effects of taking a test upon the scores of a second testing. (4) Instrumentation: Changes in calibration of the measur- ing instrument. (5) Selection: A bias resulting in differential assignment to the comparison groups. (6) Mortality: Loss of respondents from the experimental group. Sources of External Validitinhich May Limit the Study: (7) Interaction of Texting and X: In which a pretest might influence the outcome of the treatment. It is with 14Donald T. Cambell and Julian C. Stanley, Experimental Quasi-Experimentgly0esigns for Research (Chicago: Rand "CNaIlY. 1903). 90- 5-6. (8) (9) 18 this source of external validity that the researcher is most concerned. Interaction of the Selection Process and the Experi- mental Variable: This source of external invalidity may be present within the study. If, in fact, it does appear, it creates a possible source of concern for the investigator. Reactive Arrangements: Precludes generalizations con- cerning the experimental effects when the treatment is presented in non-experimental settings. Definition of Terms The following definitions are intended to define the parameters within which the researcher uses specific terms within the study. (1) (2) (3) (4) (5) Attitude: A mental or neural state of readiness to respond organized thought experience and exerting a directive and/or dynamic influence on behavior.15 Cognitive: The thought component of an attitude. Affective: The emotional component of an attitude. Conative: The behavioral component of an attitude. Belief: The acceptance of a state or condition as being true.16 Dorsey Press, 1975). 15 16 Allport, p. 810. Perry London, Beginning Psychology (Homewood, 111.: 19 (6) Opinion: An overall expression of an inferred under- lying attitude. "Opinions are viewed as verbal 'answers' that an individual gives in response to stim- ulus situations in which some general question is raised."17 (7) yglgg: A component of an attitude indicating the degree of personal acceptance. A value is the positive aspect of an attitude. (8) Message: Content and structure of what is said and the kind of appeal used. (9) Fear Message: A raising of the level of threat for non- compliance with the recommendation of a persuasive ‘ communication.18 (10) Source: The communicator of the potential attitude I change message. (11) Credibility: A receiver evaluation of the source's believability arranged in an order from high to low. (12) Attractiveness: The degree to which the receiver identifies with the source, thus enhancing his own self-esteem. 17Carl J. Hovland, Irving L. Janis and Harold H. Kelly, Communication and Persuasion (New Haven: Yale University Press, 1953). 18William J. McGuire, "The Nature of Attitude and Atti- tude Chance," in The Handbook of Social Psychology, ed. Gardner Lindsey and Elliot Aronson, 2nd ed. (Reading, Mass.: Addison-Wesley, 1969). 20 (13) Pppgp: An appraisal of the extent to which the source may administer positive or negative sanctions over the receiver. (14) Channel: The medium through which the message is presented. (15) Receiver: The individual or group, often referred to as the "target," at which the message is aimed. (16) Destination: The types of issues at which the persua- sion is aimed-long versus short effects or verbal attitude change versus gross behavioral change.19 (17) Cpgnitive Score: The knowledge component of the test instrument, compdsed of twenty-seven items. (18) Affective Score: The attitude component of the test instrument, composed of thirteen items. (19) Cognitive/Affective Score: The composite test instru- ment, consisting of forty items and measuring the total attitude orientation concerning the tested attitude object. 1 (20) Motion Picture: Within this study, the terms "film" and "motion picture" are used interchangeably. nganization of the Thesis In Chapter I, the author has included a discussion of the problem addressed within this study. Also covered in this chapter are sections which discuss: societal 19 . . Vernon Gerlach and Donald Ely, Teachin and Media (Englewood Cliffs, N.J.: Prentice-Hall, 1971), p. 289. 21 incorporation, purpose, importance, generalizability, ini- tial considerations, research questions, hypotheses, assump- tions, limitations, and a definition of relevant terms used within the study. Chapter 11 presents a review of the related literature pertinent to the experimental topic. The chapter is organ- ized in the following sequential format: Introduction, Source, Credibility, Attractiveness, Message, Fear/Threat Factors, Audience, Medium, Effect, and Summary. Chapter III presents the methodology utilized within the investigation. Included within this chapter are discus- sions covering: a description of the sample, selection of the sample. description of the treatment, instrument, pilot study, and research design, as well as the procedural format for the investigation._ In addition, the methods of analysis used concerning the data and hypotheses are presented. Chapter IV includes a discussion of the analytic procedures used with the data along with the results of the analysis. Chapter V concludes the body of the dissertation and includes discussions pertaining to: a summary of the find- ings, conclusions, implications, and recommendations for future investigations. CHAPTER II REVIEW OF RELATED LITERATURE Introduction: Major Areas of Review The organizational format for this review of related literature will follow the basic outline of Lasswell's (1948) Matrix of Persuasive Communication. The matrix defines the act of communication as an interaction of "who says what to whom through which channel with what effect."20 This statement can be organized into the following topics: who (source of message), says what (message structure), to whom (audience), through which channel (medium), with what effect (destination).21 The major emphasis of this study lies in the construc- tion of a message in an attempt to have a positive effect on audience attitUdes. Of the five topics mentioned above, source, message, and audience will receive the greatest attention in this review. The research findings in the topics discussed above provide the basis for the research questions and hypotheses. 20Gardner Lindzey and Elliot Aronson, eds., The Handbook of Social Psycholo , 2nd ed. (Reading, Mass.: Adaison-Wesley, 1969), p. . 21Ibid. 22 23 In addition, the findings will dictate the organization of the test film used within this study. It should be noted that while the matrix will provide an organizational scheme for this review, it should not be interpreted as being the only model for the communication process. Communication processes can be analyzed by the utilization of numerous principles, each yielding different sets of components. Suffice it to say that the matrix will provide the necessary organization required for this particular undertaking. Source Credibility The first area of the review involves those factors affecting audience perception of the communicator of the message. Berger (1970) reports several findings regarding this issue. He concluded that the more favorable the audi- ence evaluation of the communicator, the more the audience retained the attitude message and changed in the attitude direction. He further states that the greater the amount of factual knowledge possessed by the audience concerning the attitude topic, the more polarized was their evaluation of the source. Concerning retention of the attitude message, Berger suggests that his research findings indicate that attitude message retention is not significantly related to 24 previous factual information, pre-message attitude, post- messaqe attitude, and attitude change.22 Schweitzer (1968) investigated expertise, trustworthi- ness and style of presentation as variables affecting source credibility. His findings indicate that the three variables serve in an additive capacity to enhance the credibility of the source, whereas low credible sources delivering dull pre- sentations received low ratingscNIhis measurement scales.23 In his research on perceptions of.source credibility, . Tuppen (1972), using cluster analysis, identified five vari- ables affecting source evaluation. These included: trust- worthiness, expertise, dynamism, co-orientation, and charisma. Perceived charisma was found to correlate signif- icantly with attitude change. Tuppen's research findings further indicate that charisma was a far better predictor of change than any combination of the other variables, but that it accounted for not more than 6% of the variance in attitude change.24 In a study derived from Festinger's theory of cognitive dissonance, Bonchek (1967) concluded that when an effort is 22Wallace 6. Berger, "Some Correlates of AttitudeCMange, Retention of the Attitude Message, and Evaluation of the Com- municator," 0A1 31 (1970): 15158 (Michigan State University). 23Don A. Schweitzer, "Style of Presentation as a Credi- bility)Variable," 25 28 (1968): 4273-4274A (University of Nevada . 24Christopher J. Tuppen, "Perceptions of Source Credi- bility and Change in Attitudes," 0A1 32 (1971): 1075-1076A (University of California, Berkeley). 25 required of the audience, it is the high credible communica- tor seeking a low constant audience effort who succeeds in inducing the greatest attitude change. Also, the high cred- ible communicator seeking low levels of effort induces greater attitude change than does the low credible communi- cator seeking no audience expended effort.25 Concerning low credible message sources, Greenberg and Miller (1966) found as a result of their research that the negative effects of a low credible source can be largely obviated by delaying the identification of the source until after the message has been delivered. This delay procedure appears to operate in a manner which forces the audience to evaluate and respond to the message content on its own merits. It would appear that when the source is identified as having little or no credibility, the audience tends to be on guard and becomes immunized against any persuasive attempt. The forewarning alerts the audience that message information may be unreliable. The delaying of the identification of the source in low credible situations will only be effective, however, if the content of the message is in itself basically well organized and potentially effective.26 25Victor Bonchek, "Commitment, Communicator Credibility, and Attitude Change," EA 27 (1967): 3929-3930A (New York University. 26Bradley S. Greenberg and Gerald R. Miller," The Effects of Low Credible Sources on Message Acceptance," Speech Monographs 33, no. 2 (1966): 135. PLEASE NOTE: Page 26 is missing in numbering only as text follows. Filmed as received. 27 Johnson and Izzett (1972) investigated the effects of source identification and type of communicator. They reported that source credibility acts as an evaluative set for the audience and has an influence on their acceptance or rejection of the message. Highly credible sources gen- erally have their messages accepted as valid by participat- ing audiences. Low credible sources generally generate audience suspicion concerning the message arguments. This suspicion results in counter-arguments being generated by the audience which ultimately leads to message rejection. Concerning audience ego involvement in the message topic, Johnson and Izzett found that the phenomenon of source credibility has its greatest attitude influence effect with low ego-involved subjects. There appears to be no source credible difference in situations involving highly ego-involved audiences. Ego involvement acts alone as an influence affecting the degree to which the audience will critically evaluate the message content. Under high ego involvement, a highly critical set of evaluations generally results in nullifying source cues. Concerning the three conditions of high credibility, low credibility,(H‘unidentified source, Johnson and Izzett report, somewhat contrary to Berger's findings, that com- munication recall was not significantly affected. They sug- gest that source characteristics act to cue the audience as 28 to how the message should be received and not to what extent it should be received.27 An interesting principle presented by Goldberg (1970) suggests that high credible sources present an incentive value to the audience. This value, it is postulated, is derived from the socialization process where it is desir- able to be in agreement with authority figures. Disagree- ment with such figures often leads to undesirable consequences.28 Ostermeier (1967) investigated source credibility utilizing basically three variables: self-reference, prestige-reference, and frequency of use. He defined self- reference as a condition where the source has had first-hand experience with the message topic and reveals and stresses this involvement before the audience. ‘Prestige-reference is stated as an association between the source and an indi- vidual having first-hand involvement with the topic. Here, the source reveals and stresses this association but does not claim involvement for himself. The conclusions reached by Ostermeier lend support for the self-reference position but fail to support the 27Homer H. Johnson and Richard R. Izzett, "The Effects of Source Identification on Attitude Change as a Function of the Type of Communication," Journal of Social Psychology 86, no. 1 (1972): 81-87. 28Carlos I. Goldberg, "Attitude Change as a Function of Source Credibility, Authoritarianism, and Message Ambiguityfi' DAI 31 (1970): 464-465A (City University of New York). 29 prestige-reference position. The prestige-reference results were explained as a condition in‘which mention of the pres- tige figure did not serve to remove the state of unfamiliar- ity existing between audience and source. Therefore, frequency of use of the prestige-reference does not enhance its effectiveness or impact. Ostermeier did find, however, that frequency of use of self-reference does play a major role in message effectiveness.29 Attractiveness Source attractiveness has been extensively studied as a variable affecting communicator effectiveness. Warshofsky (1971) studied this variable and concluded that audiences having similar attitudes to those of the communicator will tend to find the communicator more attractive and that this dimension directly influences the amount of change occurring within unrelated target attitudes. It was also reported within the Warshofsky study that the greater the target atti- tude'discrepancy exhibited within a range of moderate target attitude discrepancy, the greater the resultant attitude change.30 29Terry H. Ostermeier, "Effects of Type and Frequency of Reference Upon Perceived Source Credibility and Attitude Change," Speech Monographs 34, no. 2 (1967): 142-144. 30Michael W. Warshofsky, "The Effect of a Subject- Communicator Attitudinal Similarity-Dissimilarity and Cmmmn- ication Discrepancy on Attitude Change and Demand Character- istics," DA; 32 (1971): 1198 (University of Nebraska). 30 Concerning the choice of using an attractive or expert source within the message setting, Norman (1976) reports that he found less audience agreement with the position advocated by an expert source when the source was identi- fied after the communication as opposed to before. Agree- ment with the attractive source was not influenced by a primacy or recency setting. The basic interpretation of these findings can be explained by the fact that suggesti- bility in the case of the expert source is dependent upon acceptance of supportive arguments contained within the mes- sage, whereas acceptance in the attractive situation is independent of specific arguments contained within the mes- sage. The implication of this finding, as reported by Norman, suggests that an attractive source has the capabil- ity of influencing others simply through an association with a position, without having to give supporting arguments.31 The results of the Norman study replicate the earlier find- ings of Mills and Harvey (1972), in which they also con- cluded that identification of the attractive source had no primacy-recency advantage, whereas early identification of the expert source enhanced the message effect as opposed to delayed identification.32 31Ross Norman, "When What Is Said Is Important: A Com- parison of Expert and Attractive Sources," Journal of Exper- imental Social Psychology 12, no. 3 (1976): 294-295. 32Judson Mills and John Harvey, "Opinion Change as a Function of When Information About the Communicator Is Received and Whether He Is Attractive or Expert," Journal of Personality and Social ngchology 21, no. 1 (1972): 52. 31 Horai, Naccari, and Fatoullah (1974) reported quite similar findings. They concluded that, "a source whose physical attributes become known before communication with a target tends to produce more liking and opinion agreement if the source is physically attractive rather than unattrac- 33 The Horai, et al., findings are in support of the tive." Mills and Harvey (1971) conclusions, which indicate that a message will be more persuasive if the audience perceives the source as liking them. The basic assumption, here, is that the audience expects the communicator to be more Sin- cere with them due to a perceived commonality existing between the two.34 Eagly and Chaiken (1975) investigated attractive and unattractive communicators advocating desirable and undesir- able positions. .They found that when the position advocated was basically desirable, there was no difference in communi- cator effectiveness. However, when the advocated position was relatively undesirable, the attractive communicator was far more effective in his persuasive attempt. It was also noted within this study that attractive communicators were judged more likely to advocate desirable positions, whereas unattractive communicators were expected to advocate 33Joann Horai, Nicholas Naccari, and Elliot Fatoullah, "The Effects of Expertise and Physical Attractiveness Upon Opinion Agreement and Liking," Sociometry 37, no. 4 (1974): 606. 34Judson Mills, "Effect on Opinion Change of the Com- municator's Liking for the Audience He Addressed," Psychonomic Science 25, no. 6 (1971): 335-337. 32 undesirable ones. It is interesting to note that this study also produced evidence indicating that undesirable positions were judged more likely to be advocated than desirable ones. The explanation for this;latterfinding sug- gests that students may expect authority figures to only communicate with them when there is bad news to convey.35 In a study conducted by Himmelfarb and Arzai (1974) concerning source attractiveness and choice of listening to a message, it was found that in the absence of choice, the more attractive the communicator, the greater the resultant opinion change. However, in choice situations, the opposite is true. If an individual freely chooses to expose himself to a discrepant message, the lower the attractiveness of the communicator, the greater will be the resultant opinion change. The reason for this phenomenon lies in dissonance 36 Himmelfarb and Arzai comment: theory. A consequence of choice in message exposure situations concerns what it is that the person must justify following exposure to a discrepant message and how he does it. When a person has chosen to listen to a discrepant message, he becomes responsible for the action and its effect, provided that the consequence of his action could have been foreseen with some reasonable likeli- hood. Then, he must not only justify the 35Alice H. Eagly and Shelly Chaiken, "An Attribution Analysis of the Effect of Communicator Characteristics on Opinion Change: The Case of Communicator Attractiveness," Journal of Personality and Social Psychology: 32, no. 1 (1975):TT42. 36Sammuel Himmelfarb and Daniella Arzai, "Choice and Source Attractiveness in Exposure to Discrepant Messages," Journal of Experimental Social Psychology: 10, no. 6 (1974): 517-518. 33 difference in opinion, but also his choice. He should experience dissonance and the less the credibility or attractiveness of the source, the more dissonance he has and the more he has to justify his choice. Moreover, source derogation can but increase the dissonance he experiences. The only way to reduce the dissonance that remains is to change one's opinion in the direc- tion of the message. The theory, therefore, pre- dicts an inverse relationship between source credibility or attractiveness and opinion change under choice conditions. In his research on cognitive style existing between advocate and audience, McGuckin (1967) reports three find- ings of attitude research that yield unusually consistent results. The first finding indicates that the greater the initial prestige of thecommunicator as a result of exper- tise or attractiveness, the greater the likelihood of audi- ence acceptance of the message and at least short-term opinion change. The second finding indicates that the effects generated by high or low prestige dissipate over time. This is to say that as the message elements begin to become internalized, there is a disassociation of source from message with audience response directed increasingly by message content alone. McGuckin's third finding states that retention of the attitude message is not related to variables involving high or low source prestige.38 In a study conducted by Hazen and Kiesler (1975), it was reported that to maximize audience acceptance of a 37 38Henry E. McGuckin, Jr., "The Persuasive Force of Similarity in Cognitive Style Between Advocate and Audience," Speech Monographs 23, no. 2 (1967): 145. Himmelfarb and Arzai, p. 517. 34 persuasive message, the communicator should adapt his pre- sentation to suit the particular audience addressed. This procedure dates as far back as Aristotle, who argued that the speaker needs to take into account the views of his prospective audience.39 Hazen and Kiesler further report that in speeches to black and white audiences, Stokely Carmichael "makes many of the same points. . . . The two speeches differ distinc- tively in style and persuasive appeals; however, in each instance, style and appeals are appropriate to the audience addressed.40 Message Hovland, et al., on Fear The second area of this review will analyze some of those components of the communication process concerned with message structure. To begin this section, it is perhaps I fitting to commence with the work of Carl Hovland, as much of the later research in the field is based upon his early findings or postulates. In their book entitled Communication and Persuasion (1953), Hovland, Janis, and Kelly present an interesting picture of the various processes involved in the communica- tion paradigm. They state that early in each individual's 39Michael D. Hazen and Sara 8. Kiesler, "Communication Strategies Affected by Audience Opposition, Feedback or Persuasibility," Speech Monographs 42, no. 1 (1975): 56. 4°16id., p. 67. 35 development, there exists a strong need for esteem, love, and avoidance of punishment. These component parts of the individual are satisfied by incorporating the parental set of personal values, ideals, and a variety of specific atti- tudes and beliefs. Later, the child incorporates differen- tiated attitudes as a result of peer group exposure and con- tinued family influence. These socially expected constructs are rewarded often by purely verbal forms of approval. As growth continues, the child becomes increasingly receptive to verbal symbols designating conditions of reward or pun- ishment. As adults, political, social, and ethical beliefs are fostered via a complex patterning of incentives contained symbolically within interpersonal communications and mes- sages transmitted through mass media.41 Hovland, et al., explain that when a subject is exposed to a persuasive communication, three aspects of audience responsiveness come into play. These are: 1. Attention to the varbal content of the communication, 2. Comprehension of the message of the communi- cator, 3. Afigepgggfifiipgttgfi.fignclusions advocated by These three aspects are considered to be activated under conditions which arouse audience motives. It has been sug- gested that persuasive appeals which explicitly or 41Carl 1. Hovland, Irving L. Janis, and Harold H.Kelly, Communication and Persuasion (New Haven: Yale University Press, 1953), p. 56. 42mm. p. 59. 36 implicitly predict social disapproval or some form of physi- cal danger or deprivation serve to motivate an audience to adopt the communicator's recommendations. Experimental evi- dence supports the conclusion that intensely disturbing emo- tions (fear, guilt, shame, anger, or disgust), serve as a drive state for the individual. If these elements are uti- lized within the construct of a persuasive communication, they may well serve to motivate the audience to change in the advocated direction. However, Hovland, et al., indicate that this will occur only if the tension aroused vfithin the audience is sufficiently intense as to constitute a drive state and if the condition of tension is reduced immediately after its manifestation by another message element.43 When an individual is influenced by an effective fear or threat appeal, a seqUential series of events occur. The first contact with the persuasive message usually consists, of neural content cues which define the topic for the audience. Following these cues, threat statements are delivered which begin to arouse emotional tensions. Immedi- ately following the threat elements, reassuring statements are presented indicating ways to avoid the dire consequences of the threat and thus reduce emotional tension. Reduction of emotional tension serves to reinforce the reassuring recommendation. Hovland, et al., refer to this sequence as a chain and present it symbolically as follows: 43ibid.. pp. 60-62. 37 Content cues (C) ~ Emotional riaction (E) ~ Reassuring recommendation (R). The authors indicate that there are two sets of fac- tors which influence threat/fear appeal effectiveness. The first concerns factors which account for successful arousal of emotional tension (E) and the second are those factors which reinforce the reassuring recommendations (R). Concerning the content of the persuasive message, Hovland, et al., note that the cues which act to elicit the emotional response must be familiar to the subject prior to message delivery. If a threat/fear message is to have an impact, the audience has to have had some previous learned experience with the threat such that they may undergo emo- tional arousal. To assure emotional arousal, the message content must serve to break down any rationalization on the part of the audience that they are personally exempt from the depicted threat. Message design should incorporate a personalized threat indicating that the consequences of non- conformity to recmmmmdmiactions can result in dire conse- quences to the particular individual.45 The Hovland study also supports the findings on source credibility as discussed previously within this study. Con- cerning this topic, Hovland stresses that a communicator who exaggerates the presentation of the threat situation may be faced with audience rejection of his message due to 44Hovland, et al., p. 64. 451616., pp. 66-69 38 perceptions of intent to scare. Prior exposure to prepara- tory communications which discuss the message content objec- tives tends to lessen the degree of emotional arousal elicited when the persuasive communication is finally presented. Hovland further indicates that it is essential to bring about the threat gradually so as not to overwhelm the audi- ence with anxiety. To avoid overwhelming anxiety and allow the audience the opportunity to mentally rehearse thethreat, it is suggested that the communicator utilize repeated pre- sentations of the threat, always followed by reassuring recommendations. Hovland, Janis, and Kelly investigated three types of fear appeals which they refer to as mild, moderate, and strong. Their research findings indicated that the mild fear appeal exhibited overall superiority when the objective of the communication was to create sustained preferences or attitudes.46 I In a study of fear arousing messages, Miller and Hewgill (1966) basically supported Hovland's position con- cerning emotional arousal and reassuring recommendations.47 They differ, however, from Hovland, et al., in that they postulate: 46 47Gerald R. Miller and Murray A. Hewgill, "Some Recent Research on Fear-Arousing Message Appeals," Speech Monographs 23, no. 4 (1966): 378. Ibid., pp. 73-83. 39 . that a message containing strong fear- arousing appeals will result in greater cogni- tive imbalance than will a message containing mild appeals. The grounds for this assumption are found in the fact that strong fear-arousing appeals explicitly emphasize the harmful conse- quences of failure to comply with message rec- ommendations as mild appeals do not. Message recipients exposed to strong fear appsglsshould therefore experience greater tension. Miller and Hewgill assume that greater cognitive imbalance or tension is generated by explicitly detailed, harmful consequences within the message, and that this phenomenon strongly predisposes the audience to follow the directions of the message in order to restore cognitive balance and reduce tension. A danger in this position lies in the fact that the audience may well derogate the communicator as a means of restoring cognitive balance. Miller and Hewgill's experiments with high fear mes- sages, coupled with the added variable of a high credibility source, clearly demonstrated that high fear appeals were much more effective than were mild fear, high credibility ‘messages.49 It was further reported within the Miller-Hewgill study that high fear-arousing communications presented by high credibility sources_and directed at a valued-other referent are extremely effective in eliciting the desired change. The high fear valued-other condition also produces greater 481616. , 4916id.. pp. 379-330. 40 change than does the high fear, self-referent position.50 Kraus, El-Assal, and DeFleur (1966) reported that learning from high fear-arousing appeals does not signifi- cantly differ from that of low fear-arousing appeals. Kraus, et al., indicated that within their study, those sub- jects having little experience with the subject topic exhibit the greatest attitude change after having been exposed to a high fear-arousing appeal.51 In an investigation involving social approval/social disapproval within the context of a fear-arousing message utilizing both high and low credible sources, it was reported by Powell and Miller (1967) that the social dis- approval condition resulted in greater anxiety in both the high and low credibility conditions. However, the social disapproval message resulted in greater attitude change only in the high credibility source condition. This finding appears to lend further support for the utilization of only highly credible sources within persuasive message constructs.52 50 51S. Kraus, Elaine El-Assal, and M.L. DeFleur, "Fear- Threat Appeals in Mass Communication: An Apparent Contradic- tion," Speech Monographs 33, no. 1 (1966): 28. 52Frederic A. Powell and Gerald R. Miller, "Social Approval and Disapproval Cues in Anxiety-Arousing Communica- tions," Speech Monographs 24, no. 2 (1967): 159. Ibid.. p. 386. 41 Fear/Threat Factors Conner (1972) studied fear appeals and concluded from his research that a high fear, high credibility, valued- other referent, produced the greatest attitude change, but that a high fear, low credibility, valued-other referent 53 with the exception resulted in greater behavioral change. of the low credibility finding, this research, as well as that conducted by Isaacson( 1976)54 and Rosen (1976)55, lends support for the findings reported--that high fear appeals are generally most effective in persuasive communication settings. Hass, Bagley, and Rogers (1975) investigated one ele- ment of fear appeals, that of noxiousness, and reported that this element, when incorporated within a fear appeal, facilitated attitude change. When noxiousness and the probability of an event occurring were manipulated, the var-, iable of noxiousness was found to be responsible for the resultant attitude change. The main effect of noxiousness appears to extend the full range of applicability of fear 53John J. Conner, "An Experimental Study of Fear Arous- ing Communications, Source Credibility and Communication Ref- erent in Attitude Change and Behavior," DAI 33 (1973):4654A. 54Frederick W. Isaacson, "The Effects of Defensiveness and Involvement on the Acceptance of Fear Arousing Communi- cations," 0A1 34 (1974): 6782-6783A (University of Denver). 55Theodore J. Rosen, "The Role of Fear and Danger in Persuasion Following a Fear Communication," DAI 34 (1973) (University of Wisconsin). 42 appeals. This, then, would also include message threats that do not imply physical harm.56 Lamar Reinsch (1971) investigated the use of metaphor and simile in persuasive communication constructs. He reported that utilization of the metaphor requires the audi- ence to undergo three mental processes: error, puzzlement- recoil, and resolution. Error is the immediate audience reaction upon hearing the metaphorical statement. This result is due to the individual's initial attempt to under- stand the metaphor in a literal sense. Puzzlement-recoil results as the mind rejects any literal application. Reso- lution occurs as the individual discovers similarities betweeen the topic and the metaphorical statement. Reinsch further reports that the use of the metaphor produces a state oftension(puzzlement-recoil) followed by tension relief (insight and resolution). Reinsch further states that the use of simile is analogods to, but not identical with, the metaphor. The basic difference in use is that the simile would eliminate the "error" stage, since "like" identifies the existing relationship as being non-literal. Results of the Bowers-Osborn study (1966), as reported by Reinsch, indicate that metaphors do indeed enhance atti- tude change and perhaps also add to perceived source 56Jane W. Haas, Gerrold S. Bagley, and Ronald W. Rogers, "Coping With the Energy Crisis: Effects of Fear Appeals Upon Attitudes Toward Energy Consumption," Journal of Applied Psychology 60, no. 6 (1975): 754. 43 credibility. The use of the metaphor in persuasive set- tings has been demonstrated to be more effective than the use of simile.57 James McCroskey and Walter Combs (1969) conducted an investigation into the use of "analogy" as a variable in persuasive communication. The authors stated that the anal- ogy offers either'wiform of support" or "a pattern of rea- soning" for the message construct. Whereas the simile and metaphor are limited to comparing one specific message ele- ment, the analogy affords the opportunity to point out several similar areas. It is quite possible to construct an entire message consisting of statements relating the unknown to the known through the use of analogy. McCroskey and Combs explain that because the analogy contains two sets of elements, . . . those familiar to the receiver and those unfamiliar, it helps define the meaning of a message more clearly. The analogy connects or relates these elements in the message. Assuming these stimuli produce associations in a receiver, it can be argued that a message linking these elements is more effective because it produces three sets of applications: those due to the known or unfamiliar element, and those dug to the linking of known to unknown elements. 8 A component of message design which has received sub- stantial research attention is that of the primacy-recency 57N. Lamar Reinsch, "An Investigation of the Effects of the Metaphor and Simile in Persuasive Discourse," Speech Monographs 38, no. 2 (1971): 143-145. 58James C. McCroskey and Walter H. Combs, "The Effects of the Use of Analogy on Attitude Change and Source Credi- bility," Journal of Communication 19, no.4 (1969): 338-339. p 44 effect. The order of presentation within the persuasive message has a significant influence on ultimate communica- tion effectiveness. If the purpose of the communication is to strengthen or reinforce opinions, the one-sided message is most effective if the arguments are consonant with those of the audience. It appears that two-sided messages are somewhat more effec- tive if the change agent seeks to alter the direction of a previously held opinion. The rationale for this, as reported by Rosnow (1966),suggests that arguments may inoc- ulate an audience against counter propaganda. When the message designer must incorporate negative as well as positive information, Rosnow suggests that the positIVe information be presented first (primacy). Negative arguments utilized in a primacy setting often generate an avoidance tendency on the part of the recipients. This avoidance tendency prevents or blocks the effects of later positive message elements. Rosnow further states that the presentation of positive message elements within a primacy setting, with negative components existing in a low state of salience, establishes an approach tendency within the mes- sage recipient. This strategy effectively blocks the effects of the negative elements. Concerning moderate or high interest subject matter, again, a significant primacy effect was found. Rosnow fur- ther reports, in support of previous research findings by Fisher and Lubin, Fisher, Rubenstein and Freeman, Godlberg 45 and Hovland, and Pritzker, that the greater the advocated message change, the greater the amount of change produced. When an audience is somewhat familiar with the message topic, the primacy effect should be incorporated. Con- versely, decreased familiarity with topic relates to an increased recency effect. Concerning need satisfaction on the part of the audi- ence, the message that arouses an audience need and follows this arousal with immediate satisfaction tends to elicit audience acceptance of message directives. Rosnow's findings further indicate that the first per- suasive communication received by an audience tends to freeze their position, thus leading to rejection of subse- quent counter appeals.59 The Luchins' investigation of the primacy-recency effect suggests that recency may be slightly advantageous immediately following the communication, but that the primacy effect exhibited dominance for as long as one month following initial communication.60 Weston's (1967) investigation also concerned one and two-sided message construction. His findings basically lend support to Rosnow's conclusion that two-sided messages are superior. Weston stipulates that this relative advantage 59R.L. Rosnow, "Whatever Happened to the 'Law of Pri- macy?'" Journal of Communication 16, no. 1 (1966): 13-22. 60Abraham 5. Luchins and Edith H. Luchins, "The Effects of Order of Presentation of Information and Explanatory Models," Journal of Social Psychology 80, no. 1 (1970): 69. 46 occurs under conditions of initial audience opposition to the source and when the audience is familiar with the mes- sage topic. He further cautions that the two-sided message should not include opposing arguments with which the audience is not already familiar.61 The variable of message intensity, investigated by McEwen and Greenberg (1970), offers several findings of interest for the persuasive message designer. The investi- gators experimented with the variable of "how" something is presented rather than "what" is said wtihin the message. This "how" component, referred to as "intensity," deals with the amount of deviation from evaluative neutrality in a source's statements concerning the attitude concept. McEwen- and Greenberg report that a highly credible source working with a low or non-ego-involving topic achieves greater audience results when high intensity elements are utilized. High intensity structure contributes also to source credibility. It appears that strong and more emphatic lan- guage results in increased perception of the message. Intense stimuli (loud, large, etc.) are easily identified and appear as less vague within a persuasive communication. McEwen and Greenberg state: . it has been demonstrated that certain modi- fiers are consistently rated as being of high intensity (e.g., positively, greatly, most, 61John R. Weston, "Argumentative Message Structure and Prior Familiarity as Predictors of Source Credibility and Attitude Change," EA 28 (1968): 3772A (Michigan State University). 47 definitely, extremely). While others are per- ceived as being of low intensity (e.g., perhaps, possibly, some, slightly, somewhat). 2 , Another question addressed within the McEwen-Greenberg study discusses the ratio of manipulations to total words contained within the message. Their study contained one manipulation per 9 words, while Hochever's message-(reported within the McEwen-Greenberg study) contained one manipula- tion per 16 words. The investigators concluded that increasing the frequency of the intensity ratio improves the overall results of the message. In conclusion, McEwen and Greenberg state that: . . . message intensity serves the function of orienting the receiver to certain messages or por- tions of messages, as well as affecting rggeiver evaluations of source, topic and message. In an investigation utilizing "type of persuasive message conclusions," Tubbs (1968) states: We may conclude that persuasive communications which present a complicated and unfamiliar series of arguments on impersonal topics to less intelli- gent people are more effective when the conclu- sions are stated explicitly than when the audience is left to draw its own conclusions. Tubbs provides the reader with specific definitions of implicit and explicit conclusions. The implicit conclusion 62William J. McEwen and Bradley S. Greenberg, "The Effects of Message Intensity on Receiver Evaluations of Source, Message and Topic," Journal of Communication 20, no. 4 (1970): 340-344. 63lbid., p. 349. 64Stewart L. Tubbs, "Explicit Versus Implicit Conclu- I1 32$)and5Audience Commitment, " Speech Monographs 35, no. 1 48 allows the audience the opportunity to draw its own conclu- sions based upon evidefice and arguments presented within the message, while the explicit condition exists wherein the source draws the conclusions for his audience. Tubbs concluded in support of earlier findings by Hovland and Mendell that the explicit conclusion, regardless of audience commitment, was far more effective in eliciting attitude change.65 Concerning the variable of "optimum number of message exposures," Krugman (1968) found: . . . that the average response may go up dramati- cally on the second exposure and down dramatically on the third, may go up on the second exposure and then plateau, may go up slightly on the second exposure and up slightly on the third, angfimay also go down on the second and stay down. In interpreting his results, Krugman states that there were no cases of dramatic increase of response followed by .a dramatic increase. Also, in approximately three-fourths ‘of all cases studied, the third replication produced diminished results, i.e., lower than the second response.67 Audience The third area of this literature review consists of those message factors categorized as audience variables. 65 66Herbert G. Krugman, "Processes Underlying Exposure to Advertising," American Psychologist 23, no. 4 (1968): 252. 67 Ibid., p. 18. Ibid. 49 It has been previously reported that a multitude of inter- actions come into play in any persuasive setting. Audience characteristics bear heavily on ultimate persuasive message outcome. In a study frequently cited within attitude research, Messerschmidt (1933) reports that individuals dif- fer in degree of suggestibility for different situations. Also, individuals are not considered to be generally sug- gestible or non-suggestible. A major factor in the suggest- ibility of an individual appears to be age. Aveling and Hargreaves, Binet, Cohn and Dieffenbacher, Otis and Small (cited wtihin the Messerschmidt study) indicate that as age increases, suggestibility decreases. Messerschmidt reports that suggestibility increases from age six to nine and then declines progressively after a chronological age of nine years. It appears that two important factors involved in this phenomenon are experience and training. Also, there appears to be little correlation between intelligence and the degree of suggestibility. ' 1 Within the Messerschmidt study, it was found that Opti- mal suggestibility occurred at about age seven and decreased with age from that point on.68 In a study conducted by Reymert and Kohn (1940), it was reported that suggestibility and 1.0. are virtually uncorre- lated. The investigators concluded from their researchthat: 1. There is no correlation between 1.0. and degree of suggestibility. 68Messerschmidt, pp. 422-423. 50 2. As a child's mental age increases, his suggest- ibility decreases. 3. It is not the child's rate of development, but rather his stage of development that determines suggestibility. 4. Suggestibiégty decreases as chronological age increases. Reymert and Kohn identified the five to seven-year age group as being 90% suggestible. This group is followed by the eight to ten-year age group, composed of 86% suggestflne. After a chronological age of ten years, the percentage of suggestible declines sharply. Less than 17% of individuals over a chronological age of twenty appear to be suggestible. Research findings reported by Stevenson and Siegel (1969) suggest that the dominant mode of information proc- essing for children reflects developmental differences. It appears that a child's ability to isolate and remember visual cues as opposed to auditory material is dependent upon age and grade level. Stevenson and Siegel report that: When a child begins school, his auditory percep- tual skills are superior to his visual perceptual skills, but once concentrated tuition begins, heavy demands are made on the intake of informaBion by eye in reading, hence learning is rapid. From this interpretation, the authors suggest that levels of performance for early elementary pupils would reflect an audio orientation above that exhibited for the 69Martin L. Reymert and Harold A. Kohn, "An Objective Investigation of Suggestibility," Character and Personality 9 (1940-41): 47. 70Harold W. Stevenson and Alexander Siegel, "Effects of Instructions and Age on Retention of Filmed Content," Journal of Educational Psychology 60, no. 1 (1969): 73. 51 visual. As the child progressed through the grades, a visual dominance would begin to appear.71 Concerning the topic of incidental learning, Hale, Miller and Stevenson (1968) concluded that increases in incidental learning occur between grades three and six and exhibit a decline at grade seven. The increase across the middle elementary years is attributed to newly acquired learning skills and attention to a broader range of environ- mental features. The decline in performance at grade seven is due to factors of selective attention.72 Within the Garrett study (1969), the author discusses several characteristics of young audiences (elementary and middle-school years). He indicates that: A review of the developmental literature sup- ported the assumption that young children have difficulty in dealing with abstract arguments, tend toward absolute judgments, and are not ideo- logically oriented, and thus can be expected to have more difficulty in evaluating the arguments in a perguasive message than older children and adults.7 It appears that the adolescent begins to introduce abstract variables into his evaluative processes. One such variable, salience, explains how we hold attitudes. 71 72Gordon A. Hale, Leon K. Miller, and Harold W. Stevenson, "Incidental Learning of Film Content: A Develop- mental Study," Child Development 39, no. 1 (1968): 69-76. 73Roger L. Garrett, "Developmental Changes in the Effects of Communicator Position on Attitude Change," EA 29 (1969): 4102-4103A. Ibid., p. 74. 52 Cronen's (1973) investigation involved aspects of Fishbein's Summation Theory and specifically belief sali- ence. It appears that when an individual holds an attitude, he also holds many beliefs regarding the particular atti- tude object. Also, any particular belief serves to help 'evaluate an aspect of the attitude. Cronen indicates that the strength of belief multiplied by the evaluative compon- ent constitutes one's attitude toward the object. The beliefs held by an individual about an attitude object constitutes a hierarchy. Within the hierarchy of beliefs, only the uppermost responses appear to be elicited at any given period when one attends to the attitude object. This heirarchy constitutes the individual's salient belief structure or stored information that has been prompted into the individual's consciousness. Cronen suggests that only six to nine beliefs about a particular attitude object may be salient at any one time. Also, salient beliefs determine the individual's attitude during such periods. It is further suggested that "salience" may be influ- enced by repeated exposures to various media. In other words, the position of the belief within the hierarchy may well be cued into a more favorable position by manipulation 74Ibid., p. 4103 53 75 of the exposure frequency variable. These findings on salience are supported by the research of Chaffee and Lindmer (1969)76 and Conville (1973).77 In his analysis of audience variables, McGuckin (1967) concentrated on the question of audience-advocate relation- ships of similarity and differences. He utilized the Rokeach Dogmatism Scale within his study to ascertain the ways in which attitudes are held by the subject. McGuckin reports that there are two extreme types of individuals: The highly dogmatic, as determined by a high score on the Rokeach scale views his world in a more highly polarized fashion than does the low dogmatist. He tends to see a world in which there are those who believe as he does and are therefore "for" him and those who believe differently and are therefore "against" him. He is threatened by contrary views and cannot differentiate readily among them. The low dogmatist, although he may hold atti- tudes identical to those of the high dogmatist, holds them differently. He does not tend 50> strongly to divide the world into those "for" and "against" him; he is less threatened by gontrary ideas and can differentiate among them.7 75Vernon E. Cronen, "Belief, Salience, Media Exposure, and Summation Theory," Journal of Communication 23, no. 1 (1973): 86-94. 76Steven A. Chaffee and Joseph W. Lindner, "Three Processes of Value Change Without Behavioral Change," Journal of Communication 19, no. 1 (1969): 30-40. 77Vernon E. Cronen and Richard L. Conville, "Belief, Salience, Summation Theory and the Attitude Construct," Speech Monpgraphs 40, no. 1 (1973): 26. 78Henry E. McGuckin, Jr., "The Persuasive Force of Similarity in Cognitive Style Between Advocate and Audience," Speech Monographs 23, no. 2 (1967): 146. 54 According to McGuckin, when an advocate identifies himself as being basically similar to his audience, they must readily accept him or experience dissonance. The advo- cate's identification of the existing state of similarity by itself offers the audience a form of positional support. However, it is the degree of importance attributed to the condition of similarity that determines the amount of esteem an audience will have for any message advocate. McGuckin concluded that audience esteem for a message advocate will be higher if the advocate implies a similar cognitive style. Further, an audience will exhibit greater attitude shift the more it esteems the advocate.79 Skolnick (1971) investigated the audience variable of personality as it affects persuasibility. His research, an outgrowth of earlier work by McGuire, suggests that persua- sive individuals are persons exhibiting a high need for approval. Conversely, less persuasive individuals demon-1 strated a low need for approval. The author indicated that the persuasive individual discriminates less between commun- ications of high and low quality. Also, there is less dis- 80 crimination between high and low credibility sources. Mintz (1972) also supports the finding that personality is 79 80Paul Skolnick and Richard Heslin, "Approval Depend- ence and Reactions to Bad Arguments and Low Credibility Sources," Journal of Experimental Research in Personality 5, no. 3 (1971): T99. Ibid.. D. 147. 55 a factor in determining susceptibility to attitude change.81 Rupnow (1976) investigated ego-involvement as it effects audience receptivity of persuasive message direc- tives. He concluded that the more ego-involved a subject is, concerning his basic attitude position,_the greater his resistance to persuasive attempts. Ego-involvement is reported as consisting of the degree of relevance or importance of an issue to a subject.82 In an extension of Hovland's research, Bettinghaus and Baseheart (1969) investigated audience behavioral commit- ment as a function of message directive. It had been reported that audiences, required to publicly commit them- selves to an issue while cooperating in a persuasive experi- ment, were less susceptible to counter persuasion if exposed to a "one-sided" message. Greenwald found that if an issue involving a public commitment was required within the mes- sage directive, significantly less actual behavioral change occurred. Not requiring a public commitment will generate greater audience behavioral change in the desired direction. The authors report that in presenting audiences with per- suasive messages that argue counter to audience beliefs, the 81Stanley 1. Mintz, "Attitude Change as a Function of Personality, Ego-Involvement and Transparency of the Influ- ence Attempt," DAI 32 (1972): 5451-54528. 82David E. Rupnow, "A Study of the Role of Ego-Involve- ment in Predicting the Attitude Change of an Individual Subject as a Result of Viewing One or More Persuasive Films on a Controversial Issue," DAI 36 (1976): 5186A (University of Wisconsin, Madison). T‘— 56 researcher risks total message rejection through an inoculation process.83 Brown (1972) describes the inoculation phenomenon as a basis for resistance to persuasion. He states that: . . . the development of counter-arguments by the subject in response to a defense message which serves to inoculate the subject's beleif against later attacks. This counter-arguing is stimu- lated by elements of threat contained in the defense message. Bettinghaus and Baseheart concluded, based upon their research, that when directed to commit themselves to an issue publicly, audiences not only exhibited less behavioral change, but less attitude change as well. The message design which utilizes the element of non-audience commitment coupled with a one-sided pro-attitude argument would appear to facilitate substantial attitude change.85 An audience variable of extreme importance to the message designer is that of "selective exposure." Klapper, as reported within the Paletz, Koon, Whitehead and Hagens study (1972) states that: By and large, people tend to expose them- selves to those mass communications which are in accord with their existing attitudes and inter- ests. Consciously or unconsciously, they avoid communications of opposite hue. Even Sears and Freedman in their lucid exposition of the 83Erwin P. Bettinghaus and John R. Baseheart, "Some Specific Factors Affecting Attitude Change," Journal of Communication 19, no. 3 (1969): 227-229. 84Cheryl L. Brown, "Intervening Variables in Resistance to Persuasion," DAI 32 (1972): 5421-54228. 85Bettinghaus and Baseheart, pp. 230-234. 57 analytical confusions and operational difficul- ties engendered by the term selective exposure, conclude that it still seems likely that de facto selectivity holds, as a descriptive generaliza- tion, on many occasions and for many people. By "de facto selectivity" they mean that communica- tion audiences usually share, to an extraorggnary degree, the viewpoints of the communicator. It appears that the "selective exposure" component coupled with the additional elements of "selective perception" and "selective retention" act as important reinfincement factors concerning existing attitudes.87 Wiebe (1970) also investigated the "selective exposure" phenomenon and raises several interesting points for consid- eration. He reports that since the introduction of printing in the fifteenth century, the public appetite has been for the light, superficial, trivial, scandalous, seditious, and vulgar. Even the wandering minstrel of a past era fulfilled this audience demand by delivering messages that often offended the refined taste. Wiebe proposes that the underlying cause of this audience phenomenon lies in developmental factors beginning in infancy. He stresses that the infant is preoccupied with self and attempts to levy demands on the outside world which satisfy his egocentrism. Wiebe comments: Such concepts as sharing, mutuality, reciprocal relationships. empathy, service, interaction, all of these positively valued concepts, endlessly 86David L. Paletz, Judith Koon, Elizabeth Whitehead and Richard B. Hagens, "Selective Exposure: The Potential Boom- erang Effect," Journal of Communication 22, no. 1 (1972):48. 87 Ibid., p. 49. 58 stressed in the process of socialization, turn out, on examination, to refer to rather sophisti- cated psychologically demanding processes which call for a well-developed sense of the other. These are essentially social processes which require the surrender or at least the inhibiting88 of the early deep-seated pattern of egocentrism. Wiebe stresses that media in presenting printed symbols or sounds or images, but never persons, allow the individual to revert to the early pattern of "taking without deference to the reciprocal needs of the giver (other)."89 Media affords the individual an opportunity to basic- ally return to an infantile posture in which removal of the stimulus is tantamount to its cessation of existence. states: Reality is beset with people and things that resist, react, encroach, and demand. When people are weary and frustrated and crowded, they embrace the media where people and things are ephemeral. It is characteristic of popular media content that it maximizes immediate need gratification, mini- mizes intellectual effort, and excuses the audi- ence member from acknowledging a substantial other. Wiebe further indicates that: . . .the permissiveness and phenomenal solitude that generally characterize the reception of media messages maximize the opportunity and the likeli- hood that they will be selected and perceived in ways that minister to intrapersonal impulse grati- fication. Media messages intended to change, ele- vate and refine overt interpersonal behavior will seldom succeed except where they resonate a favor- able predisposition in the receiver, and guide him 88 Audience Behavior," Public Opinion Quarterly 33, no. 4 (1970): 526-527. 891bid.. p. 527. 9°Ibid. He 90 Gerhart D. Wiebe, "Two Psychological Factors in Media 59 to face-to-face arrangements that facilitate the desired behavior.91 Medium The fourth area of this literature review summarizes some of the research findings on media effectiveness in per- suasive settings. Major emphasis will be placed upon film as it is this medium which will be utilized in the present investigation. Jacobson (1968), investigating media credibility, con- cluded that television was generally the most preferred and believed information source. Although his experimental sample indicated this overall preference, it was addition- ally noted that newspapers were actually the most widely used media.92 Keating (1972) reports several interesting findings concerning this topic. He noted that the overall preference of his experimental sample with respect to image (credibil- ity) of the communicator was live, television, written, and audio presentations. Also, favorable impressions of the communicator are enhanced within a group setting, but 91 92Harvey K. Jacobson, "The Credibil Media as Information Sources," EA 28 (1968 sity of Wisconsin). Ibid. ity of Three Mass ): 3132A (Univer- 60 greater attitude change is achieved in a solitary exposure situation.93 Allen (1975), quoting the work of Jerome Bruner, indi- cates three ways in which humans learn from experience. Enactive representation involves learning from a direct experience. Iconic representation, governed by perceptual organization, is the second type of learning. This would include various visual media such as film, still pictures, maps, graphs, etc. The third type of learning involves words or language and is referred to as symbolic representation. Concerning iconic learning, the author indicates that film facilitates cognitive changes to a greater extent than affective changes within an audience. However, Allen does offer encouragement for this investigation by reporting that some attitudinal changes did occur within his experimental sample as a result of viewing arfine-minute film.94 The experimental findings of Addis (1971) lend further support for the use of film. He concluded that film credi- bility was considerably higher and had greater overall impact than did either audio tape or written communications. This conclusion lends additional support to the findings 93John P. Keating, "Persuasive Impact, Attitudes, and Image: The Effects of Communication Media and Audience Size on Attitudes Toward a Source and Toward His Advocated Position," EAL 33 (1973): 3981-3982B. 94Ralph L. Allen, "The Effects of Viewing a Film of a Simplified Biological Food Chain Upon School Children," Education 95, no. 4 (1975): 337-341. 61 reported above by Keating, if one accepts the premise that film and television are quite similar in numerous respects?5 Worchel (1972L investigating "behavioral freedom" as an effect of film, concluded that cinema can motivate an audience to imitate behaviors demonstrated within the pre- sentation. This conclusion applies to behaviors considered aggressive as well as passive. Worchel further states that punishment for aggressive acts, incorporated wtihin the con- text of the film, may act to discourage an audience from incorporating the depicted aggressive or negative behavior?6 In a rather dated but pertinent study reported by Scollon (1956), the author indicates two findings which are relevant to the present investigation. Scollon concluded that films utilizing a communicator with a high prestige factor produce the greatest amount of attitude gain. Also, the greater the communicator-audience relationship the greater the resultant attitude shift.97 In an investigation of film and still formats, Weintraub (1973) indicates that film proved to be signifi- cantly superior. This superiority existed regardless of 95Barnett Addis, "Media Credibility: An Experimental Comparison of the Effects of Film, Audio-Tape and Written Communications on Beliefs in the Existence of Unusual Phenomena," EA; (1970): 2491A. 96Steven Worchel, "The Effects of Films on the Importance of Behavioral Freedom," Journal of Personality 40, no. 3 (1972): 433. 97Robert W. Scollon, Jr., "A Study of Some Communica- tor Variables Related to Attitude Restructuring Through Motion Picture Films," EA 17 (1957): 400. 62 subject matter, instructional objectives, grade level, sex, mental ability, or attitude.98 Concerning the comparative effects of color and black and white film on one's perception of reality, Utz (1969) concluded that color is at least as effective as black and white film presentations. Color seems to act to increase the perceived rate of action and proves to be of higher interest. It was noted that black and white films are thought to be somewhat slower in action, generally duller, and more comical than either color or mixed films.99 lsenhour (1975) investigated the effects of context and order as applied to motion pictures. He states that "two film pieces of any kind when placed together inevitably com- bine into a new concept, a new quality, arising out of the juxtaposition. Associative cues within shots serve to strengthen the assertion created by simple contiguity."100 lsenhour further reports that if a particular shot is extremely intense in comparison to other shots in the sequence, its meaning will override and counter the effect of the additional shot and it will not change. The author further clarifies this concept by stating: 98Royd L. Weintraub, "The Motion Variable in Film Pre- sentation," DAI 34 (1973): 562-563A (University of Southern California). 99Walter J. Utz, Jr., "The Comparative Effect of Color and Black and White Film Clips Upon Rated Perception of Reality," EA 29 (1969): 2664A. 100John P. lsenhour, "The Effects of Context and Orderin Film Editing," AV Communication Reveiw 23, no. 1 (1975): 75. 63 If shot A is very intense in sequences such as A, B and C, A; and B and C are neutral in intensity, then the meaning of A will remain about the same. This is because the change in meaning of a shot upon juxtaposition is inversely proportigaal to the intensity of the original meaning. The importance of the context-order principle as applied to film is illustrated by the research of Williams (1968). The author states that film, although perceived as a whole, is composed of individual elements. Altering the individual elements ultimately affects perception of the whole. Specific segments within a total film may serve to reinforce the ultimate film objective or subtract from it. Williams further states that audience interest patterns appear to follow developmental sequence. 'Interest is gen- erally low at the beginning of the film, rises by climaxes, and is highest at the end. However, the author cautions that if the relative tightness of shots is in opposition to audience interest in the subject, interest decreases. The author suggests that the use of the close-up be incorporated only to emphasize details which can be best shown with no other shot. Also, the loose-shot appears to lower audience interest. Within the reported study, it was found that scenes used for definitions, examples, and reviews were, without exception, considered to elicit high interest. Conversely, 101Ibid.. p. 76. 64 introductory sentences and clarifications resulted in a lowered interest level.102 Effect The fifth and concluding section of this literature review deals with the ultimate effect of persuasive communi- cations. Throughout the preceding four sections, various effects resulting from numerous message construct variables have been discussed. Therefore, this particular treatment will be rather brief. Greenwald (1965), in a study dealing with behavior change as a result of persuasive message exposure, raises several interesting points for consideration. He suggests that there may be three possible explanations as to why peo- ple develop specific belief behavior patterns. The first rationalizes that we might develop beliefs to justify our actions. The second poses that we may act only in accord- ance with our beliefs. The third states that individuals may not be affected by an interaction of belief and behavnnu but rather by other variables such as environmental factors. Whatever the cause, it has been demonstrated that individ- uals exhibit a high degree of consistency concerning this phenomenon. The author offers, as proof of this consist- ency, the fact that it is not uncommon for individuals, 102Robert C. Williams, "Film Shots and Expressed Interest Levels," Speech Monographs 35, no. 2 (1968):166-169. 65 under specific conditions, to undertake voluntary martyrdom.103 , In three out of four experiments, Greenwald demon- strated that when a communication advocated the importance of an action, significant belief and behavioral change occurred. The observed changes were a direct result of per- suasive aspects of the communication rather than as a result of a demand characteristic perceived within the experimental setting. The positive changes were observed to have endured for at least a two-week period. The author speculates that, contrary to the position advocated by Festinger, behavior is resistant to persuasive communications only under conditions where there is a prior commitment. Within the foUr experimental settings investi- gated by Greenwald, all demonstrated significant belief and behavioral change with the exception of one. The condition where no behavioral change occurred was attributed to the audience committing itself to an opposing position prior to message delivery. However, in this case, significant belief change was noted. Burzynoski (1977), investigating the effects of posi- tive-negative prior information on film acceptance, con- cluded that overheard positive prior information appears to 103Anthony G. Greenwald, "Behavior Change Following a Persuasive Communication," Journal of Personality 33, no. 3 (1965): 370. 104Ibid.. pp. 390-391. 66 predispose an audience to react favorably to the ensuing filmic presentation. The key variable here involves the element of overbearing.105 Cole's research (1969) involved two separate variables of attitude development. The first concerned the amount of information about the attitude concept. The second variable involved the type of bias involved in the information pre- sented. His findings suggest that intensity of attitudes increases as the amount of information presented about them and the degree of bias increases. Of the two variables, bias in information appears to have the greater effect. In addition, numerous slight bias statements are not nearly as effective as two or more strong bias statements.106 It is perhaps fitting to conclude this literature review with a report of Moldstad's research (1974). Report- ing the research conducted in 1933 by Rulon of Harvard Uni- versity, the author cites evidence of the effectiveness of film in education settings. It was determined that the utilization of films within a rote (factual learning) and eductive (general application) setting produced startling 105Michael H. Burzynoski and Dewey J. Bayer, "The Effect of Positive and Negative Prior Information on Motion Picture Appreciation," Journal of Social Psychology 101, no. 2 (1977): 215-218. 106James K. Cole, "The Effects of Amount of Information and Information Bias on Attitude," DAIIM)(1969):2395-2396B. 67 107 results. He states: The 14.8 percent superiority and 24.1 percent superiority of the text-plus-film groups over the test-only groups in immediate learning of rote and eductive items, respectively, and the 33.4 percent and 41 percent superiority of rote and eductive items, respectively, of text-plus-film groups over the text-only groups in retention of information after three and one-half months is 108 one of the most encouraging studies on film use. In closing, Moldstad quotes Tickton as he reported on the President's Commission on Instructional Technology to Improve Learning. Tickton states: The nation should increase its investments in instructional technology. thereby upgrading the quality of education, and ultimately, the quality of individual's lives and of society generally.1 9 Summary This review of related literature has attempted to bring together the most important factors and variables involved in persuasive message techniques. It was by neces- sity selective in nature as attitude research represents a voluminous accumulation of material. The present research includes studies dating as far back as 1933 and as recently as the late 1970's. In an attempt to organize this summary, the author has chosen to present this section utilizing the five 107John A. Moldstad, "Selective Review of Research Studies Showing Media Effectiveness: A Primer for Media Directors," AV Communication Review 22, no. 4 (1974): 391. 1081bid.. p. 391. 109Ibid., p. 387. 68 major sub-headings incorporated within the main body of the review. Source Summary In summarizing the research on source characteristics, it appears appropriate to categorize the findings into two main topics. The author refers to these as credibility and attractiveness. The abundance of available research sup- ports the premise that credibility of the source is not significantly affected by the degree to which an audience retains the message. Further, retention of the message is not a result of previous factual knowledge, pre-message attitude, post-message attitude, or attitude change. Charisma appears to be an important source attribute but does not account for more than 6% of the variance in attitude change. When the source of the attitude message seeks a commitment of effort from the audience, the higher the cred- ibility of the source coupled with a requested low level of commitment, the greater the audience attitude change. If the message is to be delivered by a source of known low credibility, delaying identification of the source will produce greater attitude change. This strategy forces the audience to act on the message elements alone, and by the time identification of the source takes place, attitude change will have occurred. Source credibility has its greatest effect upon an audience exhibiting low ego-involvement with the attitude 69 topic. Highly ego-involved subjects tend to nullify source cues and thus diminish the credibility effect. High credibility appears to further benefit from what is referred to as the socialization process. Within our society, it appears desirable to be in agreement with authority figures. To enhance credibility, research indicates that the message source should strive to demonstrate for his audience his association with and expertise in the attitude topic area. This strategy results in greater attitude change than association with and mention of expert or prestige refer- ences supporting the project. In addition, frequency of self-reference enhances persuasive effect. The greater the audience/source similarity, the greater the perceived source attractiveness. Attractiveness has been demonstrated as directly influencing the amount of attitude change occurring in persuasive settings. Concerning the use of an attractive or expert source in persuasive encounters, the research supports the attractive source. It appears that the attractive source has his posi- tive effect upon an audience independent of message content. While the attractive source appears most effective, the expert source has his greatest effectwhenidentified prior to the message delivery. When a position is advocated which is basically undesirable for the audience, the attractive as opposed to unattractive source has the greatest impact. However, when 70 the advocated position is basically desirable, the attrac- tiveness component reveals no difference in communication effectiveness. When an audience is given the choice of participation or non-participation in a persuasive setting, the less attractive the source, the greater the opinion change. This phenomenon has its origin in the dissonance theory. When a no-choice situation exists, attractiveness once again becomes a dominant factor in determining the degree of opinion change which will occur. Message Summary The research reported on message variables may be summarized by grouping the topics into two categories. The first major category will emphasize some of the word con- ducted by Carl Hovland and his associates. The second cate- gory will include those findings herein referred to as ' fear/threat factors. Hovland discussed the individual's need for esteem, love, and avoiding punishment in the early developmental years. He stressed how these and other needs are met via verbal cues. As the individual matures, social approval continues to be evaluated through interpersonal communica- tions and mass media. Concerning persuasive communications, Hovland discusses three aspects of audience responsiveness which affect ulti- mate message impact. These aspects or factors involve: attention, comprehension, and acceptance. The audience 71 responsiveness factors are activated when audience motives are aroused. Hovland stressed that tension arousal serving to acti- vate a drive state within the individual may be utilized by the persuasive communicator to achieve his purpose, if he can manipulate the condition so as to provide immediate ten- sion reduction after initial arousal. Hovland states that the tension reduction serves to reinforce the reassuring recommendations incorporated within the message. The author further indicates that the message designer must incorporate within his design, elements which inhibit audience rational- ization that they are personally exempt from the threat. In developing the message content, the designer must be sure to develop the threat condition gradually so as not to overwhelm the audience. This is accomplished by affording the audience the opportunity to mentally rehearse the anxiety-arousing element and by providing reassuring recom- mendations immediately following each arousal. . Hovland indicates that of the three types of fear appeals (mild, moderate, and strong), the mild fear appeal appears the most effective. A later study conducted by Miller and Hewgill lends support to many of the Hovland findings, but suggests that the strong fear condition elic- its greater change. It should be noted that in utilizing high fear appeals, the audience may derogate the source and thus void the potential effects of the communication. 72 The research on persuasive message construction seems to clearly indicate that a valued-other referent results in greater change than does a self-referent condition when incorporated within the message. Content learning is not affected by the degree of fear incorporated within a message. Also, it appears that indi- viduals are more concerned with negative sanctions than plaudits concerning social situations. Social disapproval factors account for greater attitude change mainly under conditions utilizing high credibility sources. When used within a persuasive setting, the variable "noxiousness" appears to contribute positively to attitude change. It would appear that this factor serves to heighten audience emotional arousal and point up the severity of the pending situation or message target. The use of metaphor, simile, and analogy within the persuasive message design adds to the overall effective- ness. However, research indicates that use of the metaphor, although more effective than the simile, is less effective than the analogy. The analogy offers a "form of support" and/or a "pattern of reasoning" within the message con- struct. Perhaps the greatest advantage of the analogy over the metaphor/simile is that the analogy offers the designer the opportunity to point out several similar areas, whereas the metaphor/simile are limited to the comparison of single message elements. 73 Concerning the order of presentation within the message, if both negative and positive information are to be reported, the positive information should be presented first (primacy). This, it is reported, will establish an approach tendency on the part of the recipient. This condi- tion effectively blocks the negative elements from having an effect. The primacy effect was also found to apply for condi- tions involving both moderate and high interest subject matter. Rosnow's research indicates that the first persuasive communication received by an audience seems to freeze its position, which leads to rejection of later counter persuasive appeals. I Concerning the strategy of inoculation against later counter appeals, Weston's research indicates that two-sided messages are generally superior. However, Weston cautions that in presenting the negative or opposing arguments, the designer must exhibit care so as not to introduce the audience to information with which it is not familiar. The type of conclusion used within the message con- struct was investigated by Tubbs. His findings, in support of Hovland and Mendell, indicate that the explicit conclu- sion, regardless of audience commitment, was superior to the implicit type of conclusion. 74 Audience Summary In summarizing the literature on audience variables, the findings may be grouped under a single heading of "suggestibility." It appears that optimal suggestibility occurs at a relatively early age in one's life. .Evidence seems to estdm- lish this critical period as occurring sometime between the chronological age of seven and nine years. It is generally supported that after this period, one's suggestibility decreases progressively. Early in one's academic career, the dominant mode of information processing involves an auditory orientation. Later, as greater demands are placed upon the student and as the child acquires Various reading skills, a visual dependence exhibits dominance. Incidental learning of materials also increases between grades three and six. This is due mainly to newly acquired learning skills and atten- tion to a broader range of environmental factors. A decline in incidental learning performance occurs at grade seven and is due largely to the phenomenon of selective attention. The young child has difficulty with abstract concepts and seeks solutions of absolute judgment. Also, he is not able to evaluate various arguments within a persuasive set- ting. There appears to be a natural orientation on the part of the child toward the source of persuasive message, or what is referred to as source dependency. 75 Concerning attitudes, the research suggests that indi- viduals hold beliefs about them within a definite hierarchy. Only six to nine beliefs about a particular attitude object may be salient at any one time. Salience may be manipulated by repeated exposures to a persuasive message. Personality factors involving audience-advocate simi- larity, esteem, and cognitive style affect ultimate message outcome. The available evidence indicates that persuasive individuals generally demonstrate a high need for approval. The greater this need for approval, the more attitude change one may expect. Conversely, low need for approval generally correlates with resistance to persuasion. When an audience is required to publicly commit itself to a particular issue during a persuasive encounter, a one- sided message best inoculates them against counter-arguments -or persuasion. However, research indicates that a message seeking a public commitment results in less actual behavior and attitude change than does a message that seeks no such commitment. 0f the numerous factors operating against the reception and direction of a persuasive influence attempt, selective exposure weighs heavily. The evidence suggests that indi- viduals tend to seek out those communications which lend support to their existing beliefs and attitudes. This phe- nomenon appears to be deeply rooted in our subconscious and may be.a result of a need for gratification,1ninimized intel- lectual effort, and non-involvement with a substantial other. 76 Medium Summary In summarizing the literature reported on medium vari- ables, specific attention was paid to film. It was reported that television is the most preferred and believed source of information, but that newspapers are the most widely used. Television, which the present writer considers homologous in many respect to film, is second only to live presentations concerning communicator credibility. There are at least three different ways in which humans learn from experience. These involve representations referred to as "enactive," "iconic," and "symbolic," Iconic representations involved learning from films. It appears that individuals can be motivated to repli- cate the behavior demonstrated within films. This repli- cated behavior can also include aggressive acts as well as passive ones. High prestige communicators exhibiting a close communi- cator-audience relationship produce the greatest attitude gain in films of a persuasive nature. Films utilizing continuous motion are significantly superior to those incorporating still or motioned still formats. This superiority exists regardless of subject, objectives, grade, sex, 1.0., or attitude. Color films have been demonstrated to be at least as effective as black and white presentations. It also appears that color films elicit higher audience interest. 77 Context and order, as utilized within film editing, has an important bearing on audience reception of message. Each scene has an interpretive value in proportion to the intensity of the preceding or succeeding scene. However, highly intense scenes may override this effect and thus not change in value. Audience interest in film presentations may be main- tained by manipulation of the type of film scene used. Also, scenes used for definitions, examples, and review elicit high audience interest.' Conversely, introductory scenes and clarifications act to lower audience interest. Effect Summary It has been reported within the section on the effects of persuasive communications that three possible explana- tions exist concerning the formation of belief-behavior pat- terns. Whichever pattern is ultimately responsible for this phenomenon, it appears to be a powerful one as humans demonstrate a high degree of pattern consistency. Belief-behavior changes result from persuasive elements contained within a communication and do not occur as a result of demands within the experimental setting. It appears that behavior is resistant to persuasion only when there is a commitment to an opposing attitude position prior to message delivery. Positive prior information overheard by an audience subsequent to viewing a film orients the audience to receive 78 the film in a positive manner. This phenomenon is due solely to the variable of overhearing. It was also reported that the intensity of audience attitudes increases as the amount and degree of information and bias increases. Concerning the two variables, bias has the greater effect in increasing attitude intensity. Within educational settings and used in conjunction with textbooks, film was demonstrated to have a positive audience effect. This superiority in comparison to text- book use alone exists for both factual learning and general application. CHAPTER III DESIGN OF THE STUDY This chapter will discuss the sample used within the investigation and the procedures used to design and conduct the experiment. The methods of analysis used, concerning the experimental data and hypotheses testing, will also be presented. Description and Selection of the Sample Children participating within the investigation were selected from the student population at Pinecrest Elementary School, East Lansing, Michigan. The experiment was con- ducted at three separate time intervals, commencing at 11:15 a.m., May 22, 1981, and concluding on June 8, at 12:00 noon. The Pinecrest Elementary School is located in a middle to upper-middle class residential community within walking distance of Michigan State University. Children of univer- sity faculty members are not uncommon among the student body population. The school itself is a winged, one-story, flat- roofed structure serving approximately 300 pupils. Classes range from kindergarten through fifth grade. A hearing- impaired educational program serving handicapped elementary children also operates within the building. 79 80 There exists a diversity of teaching styles at the Pinecrest School. Children in the early grades are taught in "self-contained" classes. Fourth graders learn via a "contract system" and/or operate in a "semi-departmental" environment. The fifth grade at Pinecrest is conducted in an "open space" environment utilizing "team teachers." The teaching options point to the diversification which exists within the faculty and might indicate an underlying community attitude concerning the topic of diversity itself. The student population, although consisting predomi- nantly of white Caucasians, includes representatives of minority races and nationalities. There appears to be excellent cooperation among the racial and cultural groups. The teaching staff employed at Pinecrest is highly educated. All certified teachers hold Bachelor degrees and many possess advanced graduate degrees. A number of special subject teachers also serve at Pinecrest as well as numerous paraprofessionals and students of education from Michigan State University. The test population consisted of 168 children. Fifty- three students from the fifth-grade team room participated in the study. The fourth-grade sample consisted of sixty- two students. This number represents a total of two and one-half classes, the one-half being part of a three/four split class. The remaining one-half class consisted of third graders and was considered along with an additional two third-grade self-contained classes. The additional 81 sample of fifty-three third graders completed the population for the study. Pretest On May 22, 1981, the test population gathered in the open space, fifth-grade team room at 11:15 a.m. The students were grouped together by classes, with each child having transported his or her own chair, pencil, and writing support. Cooperating classroom teachers had been briefed on pro- cedures the previous Tuesday during a meeting regarding the experiment. During the testing session, they monitored the behavior of the group and willingly cooperated when needed. After instructions were given to the students and it was assured that all understood the procedures, an agent of the researcher read the forty-item pretest aloud to the assembled group.11o Student answers indicated an intensity of agreement, disagreement, or neutrality regarding the test item questioned. Each item was read aloud twice. The pace was comfortable for the slowest students. The agent for this test was a graduate student from Michigan State Univer- sity studying deaf/blind education. She was articulate and dynamic in her role. A tape recording was made of each testing session and is available for analysis should a need arise. The agent was previously unfamiliar with the study. 11oThe pretest as well as all test instruments used within this study are found in the Appendices. 82 After the initial testing session, the computerized answer sheets were hand sorted and grouped according to grade level. They were then brought to the Michigan State University Scoring Center where the data was keypunched onto computer cards and each test was scored. Responses were compared to a predetermined correct response set. A rank order was established within each grade level, and the test scores were paired. Ranging from greatest to least score, members of ordered pairs were randomly assigned to either the treatment or the control groups. This was done for each grade level. Individual classes wtihin grades were not considered as singular units but merely added to the total grade population. Table 3.1. Population of Pretest Grade Treatment Control Total 5 26 27 53 4 31 31 62 3 26 27 53 Independent Measure Experimental Film The independent or experimental variable utilized wiUfin this study was the experimental film: a twenty-four minute, 16mm, color/sound documentary on the ecological significance of the wolf in its natural habitat. Being an experimental 83 persuasive film, certain elements within the construct of the film were manipulated in an effort to measure the 1" The model, utilized during effectiveness of the model. the production of the experimental film, consists of five basic elements. These elements include: source, message, audience, medium, and effect. It was assumed that the source of the message within the film exhibited not only an attractive attitude orienta- tion, but in addition possessed high credibility. It was further assumed that the message conveyed via the film was of a persuasive nature. Guided by Element 11 of the model concerning message content, the incorporation of a mild fear approach (see pp. 34-37) was selected for use within the experimental film. Various scenes or composite scenes were used in patterns of two throughout the film, again, directed by the model. Audio manipulations on the various sound tracks were planned for effect, as directed by the model. Language style, tone, and pace also were subject to the influence of the model. Audience factors were considered prior to the begin- ning of film production. The age group for the study was basically dictated upon generation of the main hypotheses early in the study. As the experimental population con- sisted of third through fifth-grade pupils, the attention span of the third grader had to be considered when 111See Appendix I - Model: Persuasive Message Design. 84 structuring the length of the film. It was assumed that due to a high interest content level utilized within.the film that the lower level attention span group might be able to concentrate on the subject for the additional time required to complete the film presentation. 1 The content of the persuasive message was developed after analysis of the pilot study data. Attitude and knowl- edge components were developed and carefully utilized within the construct of the film story line. The persuasive film involved an attitudinally attrac- tive American native, possessing wisdom and expertise, sUm- moned from the past to speak to the children of today. As he appears in the forest, he speaks in his native language of the perfection of nature. His message portrays the eco- logical significance of the beleaguered wolf. English trans- lations follow each Ojibwe phrase. The content of the film represents an accurate portrayal of historical, biological, ecological, and sociological facts. Wildlife footage was filmed as far north as Great Bear Lake in the Canadian Northwest Territory and as far south as Missouri. The majority of the film, however, was shot within the Algonquin Provincial Park, Ontario, Canada. Dependent Measure Test Instrument After the basic format had been established for inclu- sion within the experimental film, a panel of judges, exhib- iting expertise in the subject area, assessed a series of 85 questions that were aimed at measuring cognitive and affec- tive components presented within the film. After consider- ation and revision, an instrument emerged that attempted to measure a respondent's cognitive and affective reaction to the film experience. The test instrument utilized a Likert Scale, with the final version consisting of forty test items. Twenty-seven items were included to ascertain a "cognitive" or learning score for each respondent at each grade level tested. Simi- larly, thirteen "affective" or attitude measures were included. Originally, the affective element consisted of seven- teen items. The additional four questions were asked only of the treatment group and pertained to an evaluation of the film maker's product. These additional four questions were eliminated from the study prior to analysis of the scores in an effort to facilitate the computer program designed for sorting the randomized items for comparison and to balance the control and treatment groups.112 Respondent scores were grouped into three categories: cognitive, affective, and total. Within the analysis, each component or category was treated, resulting in three scores for each respondent. The individual scores were grouped according to treatment or control status per grade level. 112Copies of all test instruments used within the study are included as Appendices A-F. 86 Due to limits of time and expense, the test instrument was not standardized. Experimental Design The design utilized within the investigation is consid- ereda true experimental design. It assumes the format of the Pretest-Posttest Control Group Design, as illustrated within the Campbell and Stanley publication on Experimental 113 and Quasi-Experimental Designs for Research. Table 3.2. Pretest/Posttest Control Group Design _—-———_—-_-—————————— Within this design, the 0's represent the control and treatment groups. Reading from left to right across the top row of the design (above the dashed line) indicates that the treatment group was randomly assigned after receiving a pre- test. This is indicated by the assignment of the letter (R) and the numeral (1) to the treatment group (0). The (X) represents the experimental treatment (film), and the (2) assigned to the (0) represents the posttest conducted immed- iately after the treatment (X). (03) indicates a delayed posttest administered to the treatment group (0) only. 113Campbell and Stanley, p. 13. 87 The dashed line in conjunction with the (R's) indicates that a random assignment to the control or treatment group was performed, after the pretest. (04) and (05) designate the control group receiving an identical pretest (4) and posttest (5). No delayed post- test was administered to the control group. Upon scoring the pretest, subjects were matched accord- ing to their pretest composite or (total) score and randomly (R) assigned to either the treatment or control group. This procedure assured greater design precision than simply random selection alone. Utilization of the Pretest-Posttest Control Group Design eliminates or controls for the main effects of his- tory (events occurring between the first and second measure- ment beyond the experiment), maturation (processes within the respondent operating as a function of time), testing (the effects of taking a test upon the scores of a second testing), instrumentation (changes in calibrations of the measuring instrument), selection (which may affect the experimental variable), and mortality (a differential loss of respondents from the comparison groups.)114 Sources of external invalidity may involve the interac- tions of testing and the treatment. In addition, other pos- sible sources of concern, with respect to external 114lbid.. pp. 13-24. 88 invalidity, might include an interaction of selection and the treatment as well as reactive arrangements. Design Over Variables The variable matrix consists of a 2x 3 design. Inde- pendent variables consist of the experimental treatment (film) applied to the experimental group at grade levels three, four, and five. Dependent variables consist of gain scores for the cognitive achievement test and change scores for the affective attitude test items. The variable matrix is represented as: Table 3.3. Variable Matrix Grade T. C. 3 4 5 Grade: classes participating within the experiment 1.: receiving the experimental film treatment 0.: receiving the non-topic-related control film Procedures for the Investigation Preliminary Procedures Pilot Study. On February 12, 1980, a one-hour pilot study was conducted utilizing the students in grades three through five at the Pinecrest Elementary School in East Lansing, Michigan. The primary purpose of this study was to evaluate the degree of cognitive knowledge possessed by the 89 students concerning specific ecological concepts. Secondly, the survey sought to ascertain the affective disposition of the respondent concerning the experimental attitude object. The Messerschmidt study, reported in the early 1930's, sug- gests that the optimal age of suggestibility may be a 115 An extensive review of chronological age of nine years. the literature found no convincing conflicting research evi- dence to discredit the Messerschmidt findings. Based upon this study, the target population would then be composed of predominantly fourth-grade students. It was decided, that to better evaluate the Messerschmidt suggested target population, grades three and five would be included to bracket the target group. The sample involved in the pilot study consisted of twenty-five third graders representing one classroom, twenty-five fourth graders representing one classroom, and sixty-eight fifth graders representing the fifth-grade co-op class. The discrepancy in the number of children partici- pating from the fifth grade as opposed to the other classes is a result of the fifth-grade co-op being the only fifth- grade class. They therefore had to be considered as an entity.116 115Messerschmidt, pp. 422-423. 116Although the pilot study population was later uti- lized within the experimental testing, the content of the pilot study and the time delay between pilot testing and experimental testing minimized bias which might have resulted from this procedure. 90 Fifth and fourth graders read a forty-three-item survey based upon a Likert Scale. Each respondent marked on the test sheet his or her intensity of agreement concerning each question. Answer choices consisted of three options ranging from agreement to neutrality to disagreement with each item option. To facilitate the testing procedure, the forty-three- item pilot survey was read to the third-grade class by the experimenter. The pace of the questioning was slow enough to facilitate the slowest student. Each student had a copy of the test questions and marked his or her answers in the same fashion as the fourth and fifth-grade students. A panel of judges, exhibiting topic expertise, had previously assisted in the development of the test measure and response set.' Respondents' answers were compared with the "response set," and a raw score for cognitive and affec- tive questions was established for each respondent. Raw scores were converted to class percentages answering each item.117 An interpretation of the findings from the pilot study guided the selection of affective and cognitive components to be included within the experimental film. Affective and cognitive components enable the researcher to test two of 117An interpretation of the pilot study, along with accompanying test and data, may be found in AppendiceslA-C. 91 118 the three components comprising an attitude. The cona- tive or behavioral component was not measured. Teacher Meeting. On May 18, 1981, a scheduled meeting was held with the participating administrative and teaching personnel at the Pinecrest Elementary School, East Lansing, Michigan. It was during this meeting that the specific testing periods were scheduled. It was determined that the testing sequence would occur as follows: Friday, May 22, 1981, 11:15 a.m.: Pretest. Total test population. Friday, May 29, 1981, 10:45 a.m.: Treatment and Posttest. Control and Treatment groups.. Friday, June 8, 1981, 10:30 a.m.: Delayed Posttest. Treatment group only. A complete briefing on the specifics of the study fol- lowed the scheduling and coordination of the testing session times and dates. It was further agreed that all testing would take place within the open space fifth-grade classroom for treatment group involvement and in the immediately adjacent vacant classroom for control group testing. A film, of approximately the same length as the experi- mental film, was selected for use with the control group. This film, entitled The Medieval Knights, portrayed the ordeals a young boy endured as he worked toward becoming a 118Zimbardo and Ebbesen. 92 medieval knight. It was a color/sound, 16mm motion picture of twenty-two-minute length. The film subject was selected by the participating teachers from an option list of five available control films. The list of control films available for this experiment were: (l) Paddle to the Sea 28 min. (2) Legend of Johnny Appleseed 19 min. (3) Four Artists Paint One Tree 16 min. (4) A Metric America 16 min. (5) The Medieval Knights 22 min. The criteria for selecting the five control films assured that the film used would be unrelated to the subject matter presented via the experimental film. In addition, the control film had to be of a complexity level suitable for the age group participating within the study. Finally, the control film had to be of approximately the same length as the experimental film. All of the above criteria were met with the selection of The Medieval Knights as the control film. The method of scoring the various tests was discussed as well as the treatment/control group assignment procedure. In discussion related to the delayed posttest procedure of utilizing only the treatment group. it was decided that while the treatment group was involved with the delayed posttest, the entire control group would view the experi- mental film. This prevented any disruption of individual classrooms. At the testing time. children either went 93 to the treatment testing room or to the control room. Teachers were encouraged not to speak of the pending experiment with their classes. It was also requested that, as far as possible, subject matter relevant to the specific experimental topics not be presented until after the experi- mental treatment had been conducted. Collection of the Data of the Study Posttest At 10:45 a.m., May 29, 1981, the experimental film entitled Circle Song was readied for presentation to the treatment group at Pinecrest Elementary School. The original test population available for pretest procedures and assign- ment to control/treatment groups had dwindled to 157 students. Four students were absent from the fifth-grade co-op class on the date of the experimental treatment. The fourth and third-grade samples had a total of six absences. Four from the latter and two from the former level accounted for a total population loss, from pretest to posttest, of eleven students. Individual cells consisted of from twenty-two to thirty-one members. Subsequent to this experimental phase, returning absentees were omitted from further statistical consideration within the study. The treatment group gathered in the fifth-grade co-0p class at approximately 10:45 a.m. on the date of testing. Children, again as in the pretest procedures, carried their 94 Table 3.4. Population for Posttest Grade Treatment Control Total 3 22 27 49 4 29 31 60 5 EE EE 48 Total 73 + 84 = 157 own chairs to the testing center. In addition, each had his/her own No. 2 marking pencil as well as a writing support.119 After instructions for marking the computerized answer sheets were given and questions answered, the experimental 120 Two 16mm film was presented to the treatment group. projectors had been provided by the school for use in the experiment. While the treatment group was viewing the experimental film, the control group, having received identical instruc- tions, was viewing the control film in the adjoining classroom. 119The physical environment was relatively uncomfort- able for the children. The weather was hot and humid with temperatures in the mid-80°F range. The children were rather confined in their seating. Having to darken the room, the shades were drawn, adding to the uncomfortable temperature conditions already prevailing. Despite the relative discomfort, the children were cheerful, well behaved, and most manageable. 120The method for marking each answer on the computer- ized answer sheet had been developed with the pretest popu- lation. Posttest procedures were a review. 95 As in the pretest setting, cooperating classroom teachers monitored behavior in both treatment and control groups. Their mere presence seemed to preclude any deviation from acceptable behavioral norms. Agents, as in the pretest procedure, managed the treatment and control testing sessions. The two agents selected for this phase of the experiment were education majors enrolled at Michigan State University. A young man worked with the treatment group while a young woman worked with the control group. The assignment procedure was based upon an arbitrary decision. Each agent was articulate and exhibited good control over his or her group. No difficul- ties were encountered during the testing aspect of the treatment phase. The agents read each question of the forty-item test twice. The pace was comfortable for all respondents. Dur- ing the testing session involving the treatment group, a tape recording was made. As in the pretest, the tape exists as supporting evidence concerning the design validity for the proceedings. Neither agent was familiar with the study one week prior to his or her involvement. This procedure was employed to limit the degree of bias entering the design. Upon completion of the testing phase, the answer sheets were collected and the students then returned to their nor- mal schedule. No instructions were given to the treatment or control groups as to procedures to follow between 96 testing sessions. They were merely reminded that they were still participating in the experiment and one final phase remained. The posttest answer sheets were sorted into grade levels and checked individually for errors. They were then transported to the Michigan State University Scoring Center where the Information was transmitted to computer data cards. Grade levels were divided into treatment and control groups, resulting in six sets of data cards. The items utilized within the construct of the posttest as well as the delayed posttest consisted of a randomized redistribution of the pretest questions. A standard table of random numbers was used to order the questions for each CES‘C. Delayed Posttest On June 8, 1981, ten days after the posttest, the delayed posttest was administered to the treatment group at Pinecrest Elementary School. At 10:30 a.m., the children participating within the treatment group gathered in the fifth-grade co-op class. As in the past testing sessions, each participant carried his/her own chair, pencil, and writing support. The purpose of conducting the delayed posttest was to ascertain any change occurring between posttest treatment and delayed posttest results. As in the posttest, the 97 original set of test items was randomized utilizing a standard table of random numbers.121 Teachers monitored the group for behavioral deviations. The researcher conducted the delayed posttest session. As in past settings, instructions for marking the computerized answer sheets were given and questions answered. As the children utilized the same computerized answer sheet in the three test sessions. very little clarification was required. No tape recording was made of the delayed posttest session. As the purpose of the delayed posttest was to measure any delayed effect the treatment may have had upon the participating group, it was felt that any bias intro- duced by the researcher would at this time be minimal. Each test item was read twice. The pace was again based upon the comfort level of the slower students. After testing, the computerized answer sheets were collected and sorted into grade levels. During the delayed posttesting session, the control. group viewed the experimental film. This viewing served as a type of debriefing activity for the control population. Computerized answer sheets from the delayed posttest session were transported to Michigan State University Scor- ing Center where the data was keypunched onto data processing cards grouped by grade level. 121The physical environment was seasonally comfortable for the delayed posttest session. As in previous sessions, the assembled group was cheerful and cooperative. 98 The population available for the delayed posttest was identical to the posttest for statistical considerations. There were no additional absentees from the posttest popu- lation. Returning absentees missing the posttest were allowed to participate. After the session, they were elim- inated from the population being scored. Data Estimation Procedures Sorting Program Once the responses for the various grade levels and groups had been keypunched onto computer data cards for each test involvement, a computer program was written to sort the data for comparative purposes. The Fortran pro- gram developed evaluated each item in each test and com- pared individual responses across tests. Based upon an optimal score of forty points per test, respondent raw scores were generated. I 3 Each respondent answer sheet generated three total or composite scores. Mean scores for each question item were established with respect to grade level and sub-group. Affective and cognitive components were separated via the computer program, and independent scores were generated. The data that were organized via the Fortran sorting pro- gram generated the three categories upon which analysis would be conducted and evaluations made. These categories consisted of: (1) a total or composite score based upon the additive effect of combining affective and cognitive scores, (2) an affective score, and (3) a cognitive score. 99 Evaluations performed from this point utilized grade level and treatment/control status only. Individual scores were no longer required within the investigation. Having generated grade level and treatment/control group mean scores concerning cognitive/affective components as well as a total/composite score, a second computer operation was conducted. Time Series Processor (TSP) A Time Series Processor (TSP) standard program was used to conduct the data processing or computations.122 This program is commonly used in the study of econometric models. It differs from other programs utilizing analysis of variance (ANOVA) or covariance (ANCOVA) in that it incorporates a regression model. Regression analysis, as in all statistical analysis, involves inductive processes wherein generalizations are based upon observed data. The induction involved in the generalization process ultimately involves an error compon- ent. The likelihood of the occurrence of this error compon- ent is explained utilizing probability theory. Inductive inferences, stated as null hypothetical (Ho) generaliza- tions, are tested for validity by statistics and the statis- tical theory of inference. Probability theory can therefore 12‘ZStandard program: In the above case, the regression model formulated within a computer program (i.e., an algo- rithm which performs the necessary computations of statistics). 100 be used when reference is made to a particular or specific value of a statistic. A t statistic or t distribution/t test,was.used as the basic statistical device to test the validity of the main hypotheses. In each instance, the null hypothesis (Ho) was used in the statistical operation. t statistics or t tests are perhaps the most widely used tool in causal-comparative analysis. Computations conducted by the TSP program involved two phases of operation. Step number one (1) involved the estimation procedure, while step number two (2) involved hypotheses testing. Step #1. In the estimation procedure, 6 statistics were generated or calculated via application of the spe- cific TSP program component to the data. These 8 statistics equal the mean score or the mean score gain, depending on the particular equation applied or being considered. The mathematical formula used to describe 80 is as follows: E (Y/Q4=O) = B o E = expectation Y = test score Q4 = treatment group if 0, control group if 1 80 = control group mean score The formula states: the expectation of obtaining Y given the Q4 variable equaling zero is attributed to the control group mean score. 101 Hypotheses Testing Step #2. Within the second phase of the statistical computations, the 3 statistics were transformed into t statistics for the purposes of testing the appropriate hypotheses. A probabilistic value for B was determined by use of the t statistics. The formula used to calculate the t statistic is: B 4( 06 This process actually transforms the standard deviation or error and the B coefficient into the t statistic. t = t statistic B = estimator based upon data ($03) = standard deviation ofilcoefficient based upon data A total of twenty-four (24) t statistics were generated within the investigation. These values were than used to reject or accept the null hypothesis (Ho) being considered by comparing the calculated value of the experimental t statistic to a critical value for the t statistic. The critical value for t was calculated by subtracting the number of dependent variables present from the number of subjects participating within the particular component under investigation. This number then equaled the degrees of freedom operating within the investigation. Applying this number to a standard table of percentage points for 102 the t distribution and deciding as to utilization of a one- tailed or two-tailed test value, a critical value was established for t. Hypotheses and Data Analyses This investigation involved three independent varia- bles: (1) the experimental film treatment, (2) the grade level, and (3) the particular sub-group assignment (treat- ment or control). Dependent variables involved scores on the post and delayed posttests. The null hypotheses format was used in testing all major experimental hypotheses. Each hypothesis was tested at the e = .05 level. Hypothesis 1: The mean scores gain between control and treatment group for grade four is significantly greater than for grade five and grade three. Hypothesis 2: There exists a significant difference between the mean scores of treatment and control groups for each grade tested. Hypothesis 3: The mean score of grade five is significantly greater than the mean score of grade four which is sig- nificantly greater than the mean score of grade three. Each hypothesis generates three distinct statistical values. These are affective, cognitive, and a total or composite score. The three hypotheses presented constitute the major area of investigation. Because differences in the mean 103 score were the focus of analysis, t statistics were employed in the computation procedures involving all hypotheses. M): The format utilized within this study consisted of a Pretest-Posttest Control Group Design. Two classes from grades three and four along with an entire fifth grade co-op class composed the test population. After administra- tion of a pretest to the total population, paired scores from each class were randomly assigned to control or treatment group status. Treatment group respondents were tested together, as were control group members. This testing format was main- tained throughout the experiment. An experimental ecological documentary film was specially produced and used as the independent variable within the investigation. It was structured only after administration and analysis of a pilot survey. Based upon information generated by the pilot survey results and utilizing an experimental model for persuasive message design developed as a product of the literature review, the experimental film was produced. The test instrument used within the investigation was a product of information generated via the pilot survey. A panel of judges developed and assessed its effectiveness. Affective and cognitive components were measured, and a total score was generated. 104 Utilizing balanced pretreatment groups, mean change between the treatment and control group occurring after the experimental procedure could be evaluated. The method of evaluation selected was a regression model. Values for B and t were determined, and a procedure of hypotheses test- ing was employed. All hypotheses were tested at the a = .05 level. Chapter IV discusses the results of the analysis and interprets the findings. CHAPTER IV ANALYSIS OF DATA AND FINDINGS The data generated from the post and delayed post- tests were used to answer the research questions raised in Chapter 1. Testing of the various hypotheses utilized a linear regression model to analyze the data. Null (Ho) hypothesis testing under this regression equation incorpor- ated completely specified or defined qualitative variables. Within this chapter, a description of the samples and results of the analyses of the hypotheses will be presented. The Sample The original test population involved within the investigation consisted of 168 children. A total of 11 students were lost from the study prior to the experimental treatment as a result of absenteeism. Therefore, the treat- ment population consisted of 157 individuals. Of this grmux 71 participated within the treatment group, and 84 students composed the control group. This population remained con- stant for the delayed posttest session conducted ten days after the experimental treatment. Chapter 111 contains a detailed elaboration of the total population utilized withnl this investigation. 105 106 Hypotheses Six hypotheses directed the investigation. A seventh hypothesis was formulated subsequent to the initial pro- posal acceptance. As it is not properly a part of the ini- tial research proposal, it is included herein merely for consideration. Hypotheses with their null expressions are as follows: Hypothesis 1: There will be a significant difference between treatment and control groups, with respect to attitude change at each grade level tested. d 1: There will be no significant difference between treatment and control groups with respect to attitude change at each grade level tested. Hypothesis 2: There will be a significant difference between treatment and control groups with respect to cognitive change at each grade level tested. N 2: There will be no significant difference between treatment and control groups with respect to cognitive change at each grade level tested. Hypothesis 3: Fourth-grade mean gain between treatment and control groups will be significantly greater than third and fifth-grade mean gain concerning attitude change. on 3: Fourth-grade mean gain between treatment and control groups will not be significantly greater than third and fifth-grade mean gain concerning attitude change. 107 Hypothesis 6: Fifth-grade mean posttest treatment scores 0‘ with respect to cognitive learning will significantly exceed fourth-grade mean posttest treatment scores which will significantly exceed third-grade mean post- test treatment scores. 6: Fifth-grade mean posttest treatment scores with respect to cognitive learning will not signifi- cantly exceed fourth-grade mean posttest treatment scores which will not significantly exceed third-grade mean posttest treatment scores. The seventh hypothesis, developed prior to the post-I testing session, involves reactions occurring subsequent to the experimental treatment and operating up to the time of the delayed posttesting session. Hypothesis 7: There will be a significant difference in \I mean score between the posttest and-the delayed post- test concerning the treatment groUps for each grade level tested. 7: There will be no significant difference in mean score between the posttest and the delayed post- test concerning the treatment groups for each grade level tested. Results of Hypotheses Testing The first two null hypotheses can be represented by the following equation: (1.2). HD . 31 s 0 108 where 81 = change occurring as a result of the experimental variable. To accommodate the null hypothesis, the expected value of the B1 is assumed to be equal to or less than zero. Hypotheses 1 and 2 The formula used to generate the t statistic for null Hypotheses 1 and 2 then becomes: 31 = Estimator based upon the experimental data equaling the observed score. “Variance of B1 = coefficient generated via the statistical analysis; change in mean score. Results for Hypotheses 1, Grade 5 The t statistic was used to test Hypothesis 1. With the critical value of t equal to (1.645), the generated t statistic (5.258) demonstrates that a significant differ- ence exists between the two values. The test was conducted utilizing a significance level of c = .05. As the gener- ated t statistic exceeds its critical value, the null hypothesis is rejected. Results for Hypothesis 2, Grade 5 The t statistic was used to test Hypothesis 2. With the critical value of t equal of (1.655), the generated t statistic (5.385) demonstrates that a significant difference exists between the two values. The test was conducted 109 utilizing a significance level of c = .05. As the generated t statistic exceeds its critical value, the null hypothesis is rejected. The t statistic was also used to test the combined variables of Hypotheses 1 and 2. With the critical value of t equal to (1.645), the generated t statistic (6.221) demonstrates that a significant difference exists between the two values. The test was conducted utilizing a signifi- cance level of e = .05. As the generated t statistic exceeds its critical value, the null hypothesis is rejected. See Table 4.1. Results for Hypothesis 1, Grade 4 The t statistic was uSed to test Hypothesis 1. With the critical value of t equal to (1.671), the generated t statistic (.716) demonstrates that a significant difference does not exist betWeen the two values. The test was con- ducted utilizing a significance level of c = .05. As the generated t statistic does not exceed its critical value, the null hypothesis cannot be rejected. Results for Hypothesis 2, Grade 4 The t statistic was used to test Hypothesis 2. With the critical value of t equal to (1.671), the generated t statistic (3.304) demonstrates that a significant differ- ence does exist between the two values. The test was con- ducted utilizing a significance level of a = .05. As the 110 Table 4.1. Data for Ho: 1 and 2, Grade 5 B __ Estimated Standard Generated Critical ’ Coefficient Deviation t Statistic Value for t Affective Data for H3, Grade 5 30 9.1 .310 5.258 1.645 31 2.434 .463 var Bo (.096) varé1 (.214) Cognitive Data for H3, Grade 5 30 13.777 .761 5.385 1.645 31 6.131 1.138 (.582) (1.296) Composite Data for Hg’z, Grade 5 30 22.889 .923 6.221 1.645 31 8.565 1.377 (.851) (1.896) Sample = 49 Dependent variables = 2 Degrees of freedom 47 Significance level «.05 One-tailed test mean score control group change score treatment group WW> d O 111 generated t statistic exceeds its critical value, the null hypothesis is rejected. The t statistic was also used to test the combined variables of Hypotheses 1 and 2. With the critical value of t equal to (1.671), the generated t statistic (2.532) demonstrates that a significant difference exists between the two values. The test was conducted utilizing a signifi- cance level of a = .05. As the generated t statistic exceeds its critical value, the.null hypothesis is rejected. See Table 4.2. Results for Hypothesis 1, Grade 3 The t statistic was used to test Hypothesis 1. With the critical value of t equal to (1,675), the generated t statistic (1.886) demonstrates that a significant differ- ence exists between the two values. The test was conducted utilizing a significance level of c = .05. As the generated t statistic exceeds its critical value, the null hypothesis is rejected. Results for Hypothesis 2, Grade 3 The t statistic was used to test Hypothesis 2. With the critical value of t equal to (1.675), the generated t statistic (3.830) demonstrates that a significant differ- ence exists between the two values. The test was conducted utilizing a significance level of c = .05. As the generated t statistic exceeds its critical value, the null hypothesis is rejected. 112 Table 4.2. Data for Ho: 1 and 2, Grade 4 B _ Estimated Standard Generated Critical ’ Coefficient Deviation t Statistic Value for t Affective Data for H3, Grade 4 80 8.838 .492 .716 1.671 a, .506 .707 var 60 (.242) var 61 (.500) Cognitive Data for H3, Grade 4 30 11.968 .798 3.304 1.671 81 3.791 1.147 (.636) (1.317) Composite Data for Hg'z, Grade 4 80 20.807 1.180 2.532 1.671 81 4.297 1.697 (1.392) (2.880) Sample = 60 Dependent variables = 2 Degrees of freedom = 58 Significance level = “.05 One-tailed test mean score control group c 6 fig hange score treatment group 113 The t statistic was also used to test the combined variables of Hypotheses 1 and 2. With the critical value of t equal to (1.675), the generated t statistic (3.438) demon- strates that a significant difference exists between the two values. The test was conducted utilizing a significance level of a = .05. As the generated t statistic exceeds its critical value, the null hypothesis is rejected. See Table 4.3. Hypotheses 3 and 4 Null hypotheses 3 and 4 can be represented by the following equation: 0—- B4T ' B3T 5 O N I O _ < B4T BST The statistical analysis conducted on hypotheses 3 and 4 compares the specific treatment group with its control group counterpart, wherein: 80 = mean score attributed to the control group. BBT = mean score gain between control and treatment, attributed to 3rd class. 541 = mean score gain between control and treatment, attributed to 4th class. BST = mean score gain between control and treatment, attributed to 5th class. The formula used to generate the t statistic is as follows: 114 Table 4.3. Data for HO: 1 and 2, Grade 3 6 Estimated Standard Generated Critical Affective Data for H3, Grade 3 80 7.115 .464 1.886 1.675 81 1.293 .686 var 30 (.216) var B1 (.471) Cognitive Data for H3, Grade 3 80 10.923 .705 3.830 1.675 81 3.986 1.041 (.496) (1.083) Composite Data for Hg'z, Grade 3 go 18.038 1.026 3.483 1.675 31 5.279 1.516 (1.053) (2.298) Sample = 49 Dependent variables = 2 Degrees of freedom = 47 Significance level = “.05 One-tailed test 60 = mean score control group 61 = change score treatment group Coefficient Deviation t Statistic Value for t 115 A 6-13 1, _ HO.B4T‘B3TSO T - 4T 3T JVar @4T)-+var(B3T)- 2 covar (EBT, 34T) 2. _ B - é HO.B4T'85T 3 The t statistic was used to test Hypotheses 3 and 4. In addition, a composite computation was performed on com- bined or total affective and cognitive elements. With the critical value of t equal to (1.658), the generated t statistic (-.873), concerning the affective element, demon- strates that a significant difference does not exist between the two values. The test was conducted utilizing a significance level of c = .05. As the generated t statistic does not exceed its critical value, the affective element of the null hypothesis is not rejected. The cognitive element was also tested utilizing the t statistic. A critical value of t was calculated at (1.658). The generated t statistic (-.124) does not exceed its criti- cal value and the cognitive element of the null hypothesis is thus not rejected. A significance level of c = .05 was used in the study. 116 The composite computations for Hypotheses 3 and 4 also fail to reject the null hypotheses. Having a critical t ’ value of (1.658), the generated t statistic (-.447) fails to exceed its critical value at the ¢ = .05 level of significance. Results for Hypotheses 3 and 4, Grade 4 > 5 The t statistic was used to test Hypotheses 3 and 4. In addition, a composite computation was performed on com- bined or total affective and cognitive elements. With the critical value of t equal to (1.658), the generated t statistic (-2.1485), concerning the affective element, demonstrates that a significant difference does not exist between the two values. The test was conducted utilizing a significance level of c = .05. As the generated t statistic does not exceed its critical value, the affective element of the null hypothesis is not rejected. The cognitive element was also tested utilizing the t statistic. A critical value'of t was calculated at (1.658). The generated t statistic (-1.496) does not exceed its critical value, and the cognitive element of the null hypothesis is thus not rejected. A significance level of a = .05 was used in the study. The composite computations for Hypotheses 3 and 4 also fail to reject the null hypotheses. Having a critical t value of (1.658), the generated t statistic (-1.951) fails to exceed its critical value at the c = .05 level of signif- icance. See Table 4.4. 117 Table 4.4 Data for Ho: 3 and 4 . _ Estimated Standard Generated Critical 8 ' Coefficient Deviation t Statistic Value for t Affective Data for H3, 4th > 3rd or 5th 80-3 7.115 .455 30-4 8.838 .417 30-5 9.111 .447 B3T 1.293 .627 -.873 1.658 B4T .506 .600 var B3T .452 var B4T .360 var BST .445 Cognitive Data for H3, 4th > 3rd or 5th 30-3 10.923 .794 30-4 11.967 .727 30_5 . 13.777 .779 BBT 3.986 1.173 -.124 1.658 B4T 3.790 1.046 BST 6.131 1.163 -1.496 1.658 var 33T 1.376 var B4T 1.094 var BST 1.352 Composite Data for H374, 4th > 3rd or 5th §0_3 18.038 1.110 90-4 20.806 1.017 30-5 22.888 1.089 §3T 5.279 1.640 ] _ 447 1.658 941 4:297 1:463 ] -1 951 1 658 BST 8.565 1.626 ‘ ° var §3T 2.691 var §4T 2.140 var B5T 1.645 Sample = 157 Dependent variables = 6 Degrees of freedom = 151 = ‘¢.OS Significance le One-tailed test vel 118 Hypotheses 5 and 6 Null Hypotheses 5 and 6 can be represented by the following equations: I —5 IA O and N A The statistical analysis conducted on Hypotheses 5 and 6 compares the mean score of class 5 with the mean score of class 4, and the mean score of class 4 with the mean score ' of class 3. The formula used to generate the t statistic is as follows: 3 - 3 H;:85584-O T: 5 4 Jvar (65) + var (64) 2. < _ é4 ' 83 HO. 84-83-0 T- Jvar (34) + var (33) Results for Hypotheses 5 and 6, Grade 5 > 4 The t statistic was used to test Hypotheses 5 and 6. In addition, a composite computation was performed on com- bined or total affective and cognitive elements. With a critical value of t equal to (1.645), the generated t statistic (3.170), concerning the affective element, demon- strates that a significant difference exists between the two 119 values. The test was calculated utilizing a significance level of c = .05. As the generated t statistic exceeds its critical value, the affective element of the null hypothesis is rejected. The cognitive element was also tested utilizing the t statistic.~ A critical value of t was calculated at (1.645). The generated t statistic (3.112) exceeds its critical value and the cognitive element of the null hypothesis is rejected. A significance level of c = .05 was used in the study. A composite computation for Hypotheses 5 and 6 also rejected the null hypothesis. Having a critical t value of (1.645), the generated t statistic (3.489) exceeds its critical value at the a = .05 level of significance. Results for Hypotheses 5 and 6, Grade 4 > 3 The t statistic was used to test Hypotheses 5 and 6. In addition, a composite computation was performed on com- bined or total affective and cognitive elements. With a critical value of t equal to (1.645), the generated t statistic (1.349), concerning the affective element, demon- strates that a significant difference does not exist between the two values. The test was calculated utilizing a sig- nificance level of e = .05. As the generated t statistic does not exceed its critical value. the affective element of the null hypothesis is thus not rejected. The cognitive element was also tested utilizing the t statistic. A critical value of t was calculated at (1.645). 120 The generated t statistic (.637) does not exceed its critical value, and the cognitive element of the_null hypothesis is thus not rejected. A signifnxmce level of a = .05 was used in the study. A composite computation for Hypotheses 5 and 6 also failed to reject the null hypothesis. Having a critical t value of (1.645), the generated t statistic (.966) does not exceed its critical value at the c = .05 level of significance. See Table 4.5. Hypothesis 7 The seventh hypothesis can be represented by the following equation: wherein 81. the change in mean score for treatment group, is expected to be 0. The formula used to generate the t statistic for null Hypothesis 7 then becomes: 81 T:-————- Jvar 31 wherein: B = the observed mean score 1 var 81 the change in mean score 121 Table 4.5. Data for Ho: 5 and 6 é _ Estimated Standard Generated Critical ‘ Coefficient Deviation t Statistic Value for t Affective Data for H3, 5th > 4th > 3rd 83 8.409 .525 64 9.344 .457 1.349 1.645 65 11.546 .525 3.170 1.645 var B3 .276 var B4 .209 var 35 .276 Cognitive Data for H2, 5th > 4th > 3rd 83 14.090 1.006 64 15.759 .876 .637 1.645 85 19.909 1.006 3.112 1.645 var 33 1.012 var 84 .768 var 35 1.012 Composite Data for Hg's, 5th > 4th > 3rd 83 23.318 1.389 64 25.103 1.209 .967 1.645 35 31.454 1.389 3.449 1.645 var 83 1.929 var 8‘ 1.463 var 35 1.929 Sample = 73 Dependent variables = 3 Degrees of freedom = 70 One-tailed test Significance level = “.05 122 Results for Hypothesis 7 The t statistic was used to test Hypothesis 7. Three values were measured at each grade level. The critical t value (2.021) exceeded all 5th-grade generated t statistics, the statistics being affective (-.539), cognitive (-.453), and composite (-.529). The null hypothesis was thus not rejected for the fifth-grade treatment group at a signifi- cance level of a = .05. The critical t value fin:the fourth-grade treatment group was (2.005). This value exceeded all fourth-grade generated t statistics, the statistics being affective (-.286), cognitive (-.452), and composite (-.421). The null hypothesis was thus not rejected for the fourth-grade treatment group at a significance level of c = .05. The critical t value for the third-grade treatment group was (2.021). This value exceeded all third-grade generated t statistics, the statistics being affective (J16), cognitive_(-.596), and composite (-.382). The null hypothesis was thus not rejected for the third-grade treat- ment group at a significance level of a = .05. See Tables 4.6-4.8. Conclusions for Hypotheses 1 and 2 The fifth-grade treatment sample demonstrated signifi- cant change in the affective, cognitive, and composite categories tested when compared with their control counter- parts. Fourth-grade change occurred within the cognitive and composite elements, but affective change could not be 123 Table 4.6. Data for H3: Grade 5 B _ Estimated Standard Generated Critical ' Coefficient Deviation t Statistic Value for t Affective Data for H3, Grade 5 30 11.545 .298 65 -2.227 .421 -.539 2.021 var Bo ' .889 var 85 .178 Cognitive Data for H3, Grade 5 Ob 19.909 .781 35 -.500 1.104 -.453 2.021 var so .609 var 35 1.219 Composite Data for H3, Grade 5 3b 31.454 .971 35 -.727 1.373 -.529 2.021 var 3b .943 var 35 1.887 Sample = 44 Dependent variables = 2 Degrees of freedom = 42 Significance level = «.05 Two-tailed test mean score control group change score treatment group WW) 0'! O 11 ll 124 Table 4.7. Data for H3: Grade 4 3 _ Estimated Standard Generated Critical ' Coefficient Deviation t Statistic Value for t Affective Data for H3, Grade 4 60 9.344 .596 B4 -.421 .843 -.286 2.005 var 80 .355 var 4 .711 Cognitive Data for HZ, Grade 4 80 15.758 .917 O4 -.586 1.296 -.452 2.005 var B .840 ' var 84 1.681 Composite Data for H3, Grade 4 Bo 25 .103 1.390 84 -.827 1.966 -.421 2.005 var fio 1.933 var 34 3.867 Sample = 58 Dependent variables = 2 Degrees of freedom = 56 Significance level = “.05 Two-tailed test 60 = mean score control group 34 = change score treatment group 125 Table 4.8. Data for H3: Grade 3 3 _ Estimated Standard Generated Critical ' Coefficient Deviation t Statistic Value for t Affective Data for H3, Grade 3 80 8.409 .555 63 .909 .785 .116 2.021 var go .308 var 33 .616 Cognitive Data for H3, Grade 3 3b 14.90 1.025 93 .-.863 1.449 -.596 2.021 var 36 1.050 var 33 A 2.101 Composite Data for H3, Grade 3 6b 23.318 1.429 33 -.773 2.021 -.382 2.021 var 3b 2.043 var 33 4.087 Sample = 44 Dependent variables = 2 Degrees of freedom = 42 Significance level = «.05 Two-tailed test 8 3° mean score control group 3 change score treatment group 126 verified by the components' generated t statistic. Sig- nificant change occurred within the third-grade treatment sample in affective, cognitive, and composite categories. Conclusions for Hypotheses 3 and 4 The fourth-grade treatment group's mean gain was not significantly greater than its counterparts' in the third and fifth grades. This finding applied to the three cate- gories tested: affective, cognitive, and composite. Conclusions for Hypotheses 5 and 6 Concerning a comparison of mean posttest scores, the fifth-grade treatment sample obtained a significantly higher mean score than the fourth grade comparison group in all categories tested. When compared with the third-grade treatment sample, the fourth-grade treatment group posttest scores failed to demonstrate statistical significance at an a = .05 level. Conclusions for Hypothesis 7 The statistical measure of the post-delayed-post data supports a finding of no significant change in mean treat- ment score having occurred. This finding is applicable to all categories tested at each grade level. Summary The test data generated as a result of participation within a treatment/control group design experiment were used to test a total of seven hypotheses. The t statistic was 127 the method of analysis used, with a level of significance established at the c = .05 level. A total of thirty statistical findings concerning affective, cognitive, and .composite categories were generated. Expressed as null hypotheses, nineteen were not rejected whereas eleven were. However, Hypothesis 7 generated nine null hypothetical statements, and no significant difference between post and delayed posttest scores were found. In this instance, failure to reject the nine null hypothetical statements lends further credence to the findings. Table 4. 128 9. Descriptive Information Treatment Posttest Delayed Grade Sample Mean* SD Mean SD 3 22 23.318 1.389 22.545 2.021 4 29 25.103 1.209 24.276 1.966 5 EE 31.454 1.389 30.727 1.373 73 Control 3 26 18.038 1.026 4 31 20.807 1.180 5 E1 22.889 .923 84 *Mean represents the composite elements of the measure Table 4.10. Summary of the Findings Grade Affective Cognitive Composite Null Hypotheses 1-2 3 rejected rejected rejected 4 not rejected rejected rejected 5 rejected rejected rejected Null Hypotheses 3-4 not rejected not rejected not rejected not rejected 3 not rejected 4 . not rejected 5 Null Hypotheses 5-6 not rejected not rejected 3 not rejected 4 rejected 5 rejected rejected Null Hypothesis 7 3 not rejected not rejected not rejected 4 not rejected 5 not rejected not rejected not rejected not rejected not rejected CHAPTER V SUMMARY, CONCLUSIONS AND RECOMMENDATIONS The study was conducted to measure the effectiveness of documentary film as an attitude influence device on elemen- tary school children. Children participating within the investigation were enrolled in grades 3, 4, and 5. This population was suggested by information obtained through a review of the pertinent literature. A model for persuasive message design was developed and guided the selection of various components incorporated within the experimental film. Utilizing a treatment control group experimental design, measurements were made of the effectiveness of the experimental film with respect to affective, cognitive, and composite elements. Assuming the critical target sample to be the fourth- grade treatment group, specific hypotheses were presented in an attempt to measure any uniqueness which might exist. A Likert Scale was developed with the assistance of a panel of judges and was used as the test instrument within the experiment. Test items were randomly redistributed for post and delayed posttesting sessions. 13D 131 Prior to this investigation, little appears to have been accomplished with respect to attitude/film experi- mental research concerning elementary school children. This may be due partially to the high costs involved in producing a specific experimental product. Writingvfithnlthe Handbook of Social Psychology, McGuire (1969) reports: Concerning the persuasive impact of mass media, there is little evidence of attitude change and much less change in gross behavior. Some of the results make it appear that mass media campaigns may even have the reverse of the intended persuasive impact. Since communication campaigns designed to change attitudes and gross behavior via the mass media represent an invest- ment of $20,000,000,000 per year in the U.S. alone, which probably makes it our third largest investment of national resources, those who are directly or indirectly engaged in this entennfise are loath to accept these negative findings at face value.123 The experimental film developed specifically for this investigation utilized interpreted results from the pilot study to guide its production. Audience factors were analyzed, and these, in turn, controlled the length, inten- sity, and complexity of the experimental film. Three exper- imental elements were measured to ascertain the effectiwymss of the film's persuasive attempt. Affective, cognitive, and composite scores were examined via a regression model utilizing t statistic comparisons. A review of the related literature generated the topic components used in the experimental model of persuasive 123McGuire 132 message design. Laswell's “Matrix of Persuasive Communica- tion" guided the review of the literature and the subsequent 124 This matrix involves components of model development. Source (the who of the message), Message (the what of the message), Audience (the whom of the message), Medium (the how of the message), and Effect (the result of the mes- 125 These components were then investigated as sepa- sage). rate entities via the literature review. This procedure resulted in the development of "elements of persuasion" suitable for incorporation within the construct of the experimental film. Organization of the elements of persua- sion resulted in a new model which subsequently directed this investigation. The model served to guide the topic format included within the construct of the pilot survey. Results of the survey generated attitude and learning com- ponents to be measured during the investigation. As in the pilot survey, the model also directed the utilization of the attitude/knowledge components within the experimental film. Attitude/knowledge components incorporated within the construct of the experimental film were used in developing a Likert Scale. This instrument attempted to assess the effectiveness of the film on the target population. A panel of judges exhibiting topic expertise assisted in the development of the instrument and subjectively certified it 124 125 Lindzey and Aronson, p. 172. Ibid. 133 to represent an accurate assessment device for measuring those attitude and knowledge components under investigation. The population considered for utilization within the experimental study was determined by the findings of the 126 First reported in 1933. the find- Messerschmidt Study. ings of the study suggested that the optimal age of suggest- ibility occurs at a chronological age of nine years. As no research evidence could be found to discount the Messer- schmidt findings, the researcher utilized this age group within the study. In an effort to isolate a "critical" age group (or, as in the case of this investigation, a grade level), bracketing was utilized within the design construct. This procedure merely involves including grade levels on each side of the target grade within all elements of the study. Participating students were enrolled in the Pinecrest Elementary School located in East Lansing, Mihcigan. Grades three, four, and five were included in the investigation. Two classes from each of the third and fourth grades and a composite group of fifth graders composed the test popula- tion. The fifth-grade composite group was composed of the entire fifth-grade population equaling approximately three classes. After a meeting with the cooperating administrator and faculty, during which time the schedule and procedures for 126Messerschmidt, pp. 422-423. 134 the experiment were considered, the experimental aspect of the investigation was conducted. A pretest was administered to the total test popula- tion, and scores were paired by grade level. Random assign- ment to the treatment or control groups was performed for each grade level. Six samples were generated for the experimental procedures. The experimental procedure involved the projection of the test film, "Circle Song" (24 minutes, color), and subse- quent testing of affective disposition and cognitive learn- ing resulting from exposure to the film. A balanced control group viewed a film of similar length but unrelated to the topic under investigation. A posttest was also administered to this group. Due to the balance that existed between the treatment and control groups as a result of the matched score pairing process, results of group scores could be statistically compared. Any change occurring in the treatment group would be attributed to the experimental variable. A delayed posttest followed the main experimental procedure by a ten-day period. Only the treatment group sample was tested during this session. The delayed post- test was designed to analyze the "decay" or "sleeper effect" phenomenon commonly associated wtih persuasive or influence attempts. It involved a randomized redistribution of the original pre/posttest measures. 135 Hypotheses testing utilized a comparison of each equation's generated t statistic with its critical value. The critical value of t for any one equation was generated by subtracting the number of dependent variables from the population sample. This number, the degrees of freedom operating within the particular experimental components, was then compared to its corresponding value on a standard table of percentage points of the t distribution. The correspond- ing value then equals the critical value for t. If the generated t exceeds its critical value, the null hypotheti- cal statement concerning that equation is rejected. By rejecting the null, we thus accept the premise of its counterpart, the hypothesis. The results of the hypotheses testing procedure are as follows: H0_1: There will be no significant difference between treatment and control groups wtih respect to attitude change at each grade level tested. Concerning the 1 third and fifth grades, the hypothesis was rejected. The fourth grade failed to reject the hypothesis. ”0-2: There will be no significant difference between treatment and control groups with respect to cognitive change at each grade level tested. The hypothesis was rejected for all grade levels tested. “0-3: Fourth-grade mean gain between treatment and con- trol will not be significantly greater than third and 0-4‘ 0-5‘ 0-6‘ 0—7‘ 136 fifth-grade mean gain concerning attitude change. The hypothesis was not rejected. Fourth-grade mean gain between treatment and con- trol will not be significantly greater than third and fifth-grade mean gain concerning cognitive learning. The hypothesis was not rejected. Fifth-grade mean posttest treatment scores with respect to attitude will not significantly exceed fourth-grade mean posttest treatment scores, which will not significantly exceed third-grade mean posttest treatment scores. Concerning the fifth grade, the hypothesis is rejected. The fourth-grade sample data failed to reject the hypothesis. Fifth-grade mean posttest treatment scores with respect to cognitive learning will not significantly exceed fourth-grade mean posttest treatment scores, which will not significantly exceed third-grade mean posttest treatment scores. Concerning the fifth grade, the hypothesis is rejected. The fourth-grade sample data failed to reject the hypothesis. There will be no significant difference in mean score between the posttest and the delayed posttest concerning the treatment group for each grade level tested. The hypothesis was not rejected. 137 Conclusions Testing of the hypotheses resulted in the following conclusions. Since Hypotheses 1 was rejected for the third and fifth-grade samples, it is concluded that change did occur within the treatment sample. Further, this significant change was not found within.the control group counterpart. It would appear that the experimental film was the variable responsible for the observed change. Since the observed change was in a positive direction, in line with the researcher's design, it is concluded that this documentary film can affect viewer attitudes to a sig- nificant degree (c = .05), when utilized in conjunction with similar population considerations. Hypothesis 2 was also rejected for the third and fifth- grade samples. As in Hypothesis 1, the variable responsible for the significant change in treatment scores is the exper- imental film. The significantly greater cognitive score of the treatment group represents the measure of the learning component within the experiment. It is concluded that substantial learning took place as a result of viewing the experimental film. Therefore, film may be effective when used as a teaching device within a population similar to the experimental population. Composite scores, representing a total attitude orien- tation toward the experimental attitude test object, also support the conclusion that the experimental film was 138 responsible for the positive change which occurred. These composite findings were derived by combining affective and cognitive test items. This procedure attempted to measure two of the components comprising an attitude. The conative or behavioral component was not controlled or measured within the limits of this investigation. It is important to note that the fourth-grade treat- ment sample was relatively unaffected by the attitude mes- sage presented through the film. Although substantial cog- nitive change did occur between the treatment and control groups, the difference proved insignificant (c = .05) when compared with the third and fifth-grade gains. Utilizing subjective interpretations of the pilot sur- vey findings in addition to the statistical conclusions relative to H0 1 and 2, it might be concluded that factors extraneous to the design of the study were operating to produce the disparity in gain score observed in the fourth- grade treatment sample. It might be further speculated that this difference is developmental in nature. The findings of significant change (at = .05) occurring at the fifth and third levels would appear to support this developmental explanation. The evidence for developmental functions operating to affect the study gains further sup- port when it is noted that significant (¢ = .05) gain in the cognitive area did occur. As the learning scores for the fourth-grade treatment group exhibit significance (a = .05) internally (treatment/control) but fail to exhibit external 139 significance (treatment/treatment), it is suggested that exterior factors operating to suppress attitude influence operated to a lesser, but still evident, degree to suppress the cognitive score as well. This phenomenon perhaps mani- fests itself in the interpretation of Hypotheses 3 and 4. It was hypothesized that the fourth-grade target popu- lation, being in the critical age and grade range, would exhibit positive change significantly greater (¢ = .05) than its third and fifth-grade counterparts. The generated t statistics failed to support this hypothetical position, and the null hypotheses were not rejected. The conclusion drawn from the analysis of Hypotheses 3 and 4 suggests that the fourth-grade sample failed to exhibit a significantly (c = .05) greater attitude susceptibility on the experi- mental measure when compared with its third and fifth-grade counterparts. Factors operating to influence the affective element could also explain the degree of cognitive suppression observed. Hypotheses 5 and 6 were rejected for the fifth grade. Therefore, the mean score for the fifth-grade posttest was significantly («== .05) greater than the fourth grade's corresponding score. This finding is applicable across the three categories examined (affective, cognitive, composite). The fourth-grade scores did not demonstrate significance (c = .05) when compared with the third-grade sample. This lack of significance was evident in all categories tested (affective, cognitive, composite). This finding seems to 140 support the earlier conclusions drawn as a result of examining Hypotheses 1-4. Hypothesis 7 sought a determination of the degree of change occurring within the test sample as a function of time operating after the initial exposure to the experi- mental treatment. The null form of the hypothesis was not rejected. The conclusion drawn from this result indicates that after a substantial period of time (ten days), the measured change, which was recorded within the treatment population, had not altered significantly (c = .05). No "sleeper effect" was observed. Conversely, no substantial decay had taken place. It is important to note that the experimental film "Circle Song" appears to have signifi- cantly altered sample attitudes, and that this recorded change remained significantly (at = .05) intense after a time interval of ten days. In addition, there appeared no sig- nificant change in cognitive or composite scores for each grade level tested. The non-significant fourth-grade affective treatment score underwent no significant (¢ = .05) change from post to delayed posttesting. Discussion The experimental findings generated within this investigation revealed that the "model" for persuasive mes- sage design developed for this study may have contributed to the ultimate positive results attributed to the experi- mental film. The elements contained within the model 141 directed the organizational format of the experimental film. Perhaps the most important finding of this study is the empirical evidence supporting the conclusion that the experimental film was effective in changing attitudes of third and fifth-grade elementary pupils. Substantial learn- ing also resulted at each grade level as a result of viewing the experimental film. The study concluded that fourth-grade students were not significantly affected by the film concerning attitude dis- position. It is speculated that specific factors, external to the study and unique to the fourth-grade population, may have operated to produce the observed discrepancy. The researcher further speculates that these factors may be of a developmental nature. I It was expected that a hierarchy of score intensity would exist in reporting the grade level findings. This did not prove to be the case when evaluated in accordance with a significance level of e = .05. Although the fifth-grade mean score was significnatly greater than the fourth-grade mean score in all areas tested, the fourth grade failed to demonstrate significance over the third-grade scores. This finding was applicable for all categories measured. The delayed posttest measure indicated that no signifi- cant change had occurred regarding the elements tested at each grade level ten days after initial exposure to the experimental treatment. This finding would appear to 142 support the effectiveness of, and establish a level of credibility concerning, the experimental film. The experi- mental factors stimulating change within the film were of such apparent intensity that decay was not significantly apparent at the time of delayed posttesting. Implications The following implications, based upon the experimental conclusions of the study, are presented for consideration. Contrary to the existing negative research findings concerning the effectiveness of film as an attitude influ- ence device, the present investigation concluded that the experimental film used as the treatment variable was success- ful in altering expressed attitudes with respect to the attitude object. In addition, the attitude shift maintained its original generated intensity when measured against a significance level of a = .05 at a time interval ten days beyond the experimental exposure. It is suggested, as a result of this study, that atti- tude influence procedures and strategies can be incorporated within the construct of motion picture productions such that the resultant product would be a dynamic attitude influence device. It would further appear that strategies employed in the construct of the film could be applied to any organiza- tional system seeking to instill expressed attitude change into its operation or procedures. An additional implication of this study is that learn- ing can be significantly enhanced through specific 143 manipulations of the medium. Gain scores at each level demonstrated that substantial learning had occurred as a result of viewing the experimental film. It is suggested that this phenomenon may be the result of developing a target-specific film. The film is target specific in that its component parts were developed only after a pilot sur- vey was administered to assess the attitude and knowledge held by the respondent regarding the attitude object. It was therefore possible for the researcher to develop a film which included new topic and attitude information. The implication of this finding is that in persuasive message design settings, the greater the amount of knowledge held by the designer concerning the target population, the greater the design precision, thus enhancing the probability of success. A final implication concerning the experimental dis- crepancy noted with respect to the fourth-grade affective score might be explained as simply additional supporting evidence for continued acceptance of the findings of the Messerschmidt study.127 If, as reported, optimal suggestibility concerning attitude influence occurs at or about a chronological age of nine years, then it could be argued that the phenomenon witnessed was merely the end of the optimal suggestibility range, manifest in the findings of grade population scores. 1271bid. 144 As indicated in the literature review, young children especially susceptible'to persuasive messages tend to accept the views of authority figures without questioning the content of the message. This is due primarily to their inability to deal with abstract/complex message structure.128 Continuing this argument, it could be speculated that the fourth grade, exceeding a critical level of development, was merely unwilling to accept the advocated position at face value and lacked sophistication or refinement in developing abstract reasoning ability to deal with the points presented within the persuasive argument. The sig- nificant (3 = .05) increase detected between the fourth and fifth-grade sample might then be explained by the dif- ference in level of abstract reasoning ability possessed by the two groups. Recommendations for Further Investigation The data obtained within the experimental procedure suggest other areas in need of investigation. The findings indicate that documentary film is an effective attitude influence device when used in conjunction with specific per- suasive techniques incorporated within the design construct. As this finding is contrary to much of the existing research evidence, additional investigations should be conducted in an attempt to address this issue. 128Garrett 145 Also, as the research was based on establishing the degree of effectiveness of the experimental film, attitude components were measured collectively. Research is needed to isolate those elements of persuasion having greatest impact on target film audiences. Elements having negative effect or of measured low intensity should be isolated and eliminated from experimental designs when feasible. The Laswell Model offers a useful format for future 129 Topics of Source. Message. Audience. investigations. Medium, and Effect were incorporated within the construct of the experimental film. Intensity of effect of these various components of the message design paradigm needs greater investigation. The experimental finding concerning the fourth-grade affective score poses interesting questions for future research. As the score was not significantly changed as a result of the experimental variable, developmental factors were suggested as a possible explanation. This question needs further investigation to explain its significance. In conclusion, and again pertaining to results of the fourth-grade affective sample, it would appear that the nature of this apparently unique group also needs greater study. The question of why and how they differ from their third and fifth-grade counterparts needs further explanations. 129Lindzey and Aronson BIBLIOGRAPHY BIBLIOGRAPHY Addis, Barnett. "Media Credibility: An Experimental Compar- ison of the Effects of Film, Audio Tape and Written Communications on Beliefs in the Existence of Unusual Phenomena." EAL 31 (1970): 2491A (University of Oklahoma). Alexander, Sue C. "An Axiomatic Representation of Festin- ger's Theory of Cognitive Dissonance: An End and a Beginning." DAI 31 (1971): 3634A. Allen, Ralph L. "The Effects of Viewing a Film of a Simpli- fied Biological Food Chain Upon School Children." Education 95 (1975): 337-344. Allport, Gordon W. "Attitudes." In A Handbook of Social Psychology. Edited by Gardner Lindsey and Elliot Aronson, nd ed. Reading, Mass.: Addison-Wesley, 1969. Andrews, Charles R. "The Effect of Short-Term Interruptions Upon Factual Acquisition from Motion Picture Films Among Intermediate Elementary School Pupils." EA; 32 (1971): 109A (University of Oklahoma). Barber, Theodore Xenophon, and Calverley, David Smith. "Hypnotizability, Suggestibility, and Personality: A Study With the Leary Interpersonal Check List." British Journal of Social and Clinical Psychology 3 (1964): 149-150. Berger, Wallace G. "Some Correlates of Attitude Change, Retention of the Attitude Message, and Evaluation of the Communicator." EAL 31 (1970): 15158 (Michigan State University). Bettinghaus, Erwin P., and Baseheart, John R. "Some Specific Factors Affecting Attitude Change." Journal of Communication 19, no. 3 (1969): 227-238. Bonchek, Victor. "Commitment, Communicator Credibility and Attitude Change." EA 27 (1967): 3929-393OA (New York University). Bonoma, Thomas V., and Tedeschi, James T. "Some Effects of Source Behavior on Target's Compliance to Threats." Behavioral Science 18, no. 1 (1973): 34-41. 146 147 Breglio, Vincent J. "The Effects of Own-Attitude, Ego- Involvement and Open/Closed Mindedness on the Judgment of a Persuasive Communication." DAI 33 (1973): 60568. Bressler, Joanna M. "Effects of Negative Communicator and Attitude Similarity Upon Attitude Change." DAI 31 (1971): 75668 (New York University). Brooks, W.D. "Effects of a Persuasive Message Upon Atti- ‘ tudes: A Methodological Comparison of an Offset Before- After Design With a Pre-Test--Post-Test Design." Journal of Communications 16, no. 3 (1966): 180-188. Brown, Cheryl L. "Intervening Variables in Resistance to Persuasion." DAI 32 (1972): 5421-54228 (University of California, Davis). Brown, Walter A., Corriveau, Donald P., and Monti, Peter M. "Anger Arousal by a Motion Picture: A Methodological Note." American Journal of Psychology 134, no. 8 (1977): 930-9311 Bruha, John J. "Stimulus-Approach Tendencies of Learners as a Factor in Evaluation of Instructional Films." DAI 31 (1971): 3406A. ‘ Bruvold, William H. "Attitudes Toward Science and Accom- panying Beliefs." Journal of Social Psychology 94, no. 2 (1974): 269-274. ~ Bryan, William L. "Toward a Viable Environmental Movement." Journal of Applied Behavioral Science 10, no. 3 (1974): 387-401? - Burzynski, Michael H., and Bayer, Dewey J. "The Effect of Positive and Negative Prior Information on Motion Pic- ture Appreciation." Journal of Social Psychology 101, no. 2 (1977): 215-218. Calder, Bobby J. “The Effects of Cognitive Constraints on Attitudes." DAI 31 (1971): 69228. Campbel, Donald T., and Stanley, Julian C. Experimental and Quasi-Experimental Designs for Research. Chicago: Rand McNally, 1963. Carson, Rachel Louise. Silent Spring. Boston: Houghton Mifflin, 1962. Capon, Noel, and Hulber, James. "The Sleeper Effect--An Awakening." Public Opinion Quarterly: 37, no. 3 (1973): 333-358. 148 Chaffee, Steven A., and Lindmer, Joseph W. "Three Processes of Value Change Without Behavioral Change." Journal of Communication 19, no. 1 (1969): 30-40. Cohen, Stanley H. "Models of Attitude Change in the Passive Communication Paradigm: Information Processing, Social Judgment, Dissonance, Balance and Congruity." EAL 32 (1972): 72888. Cole, James K. "The Effects of Amount of Information and Information Bias on Attitude.‘ DAI 30 (1969): 2395-23968 (University of Kentucky). Conner, John J. "An Experimental Study of Fear Arousing Communications, Source Credibility and Communication Referent in Attitude Change and Behavior.“ EAL 33 (1973): 4564A (Louisiana State University). Cronen, Vernon E. "Belief, Salience, Media Exposure, and Summation Theory.“ Journal of Communication 23, no. 1 (1973): 86-94. Cronen, Vernon E., and Conville, Richard L. "Belief Salience, Summation Theory and the Attitude Construct." Speech Monogpaphs 40, no. 1 (1973): 17-26. Cronkhite. G. "Autonomic Correlates of Dissonance and Attitude Change." Speech Monographs 33, no. 4 (1966): 392-399. Domyhan, Roger A. "The Effects of a Non-Persuasive Film, a Specially Designed Persuasive Criticque and a Non- Persuasive Group Discussion on Attitude and Retention of Information." EAL 33 (1973): 6586A. Doran, Rodney L., Green, Richard A., and McIntyre, Patrick J. "Construction of Test Items to be Presented Via Motion Picture Film" Audiovisual Instruction 19, no. 2 (1974): 51-56. Eagly, Alice H., and Chaiken, Shelly. "An Attribution Ana- lysis of the Effect of Communicator Characteristics on Opinion Change: The Case of Communicator Attractive- ness.“ Journal of Personality and Social Psychology 32, no. 1 11975): 136-144. Emery, Edwin, Ault, Philop A., and Agee, Warren K. Intro- duction to Mass Communications. New York: Dodd, Mead and Company, 1973. Eyestone, Merle L. "A Comparison of the Effectiveness of Bulletin, Film, and Lecture With and Without Discus- sion, in Presenting Research Information." EA 27 (1966): 922-9238. 149 Fearing, Franklin. "Motion Pictures as a Medium of Instruction and Communication." .University of Califor- nia Publications in Culture and Society 2, no. 3 (1950): 101-102. Fielder, John F. "The Relationship Between Intensity of Belief and Level of Cognitive-Complexity." DAI 31 (1970): 2175A (Michigan State University). Fitzsimmons, Stephan J., and Osburn, Hobart G. "The Impact of Social Issues and Public Affairs Television Docu- mentaries." Public Opinion Quarterly 32 (1968): 380-397. Foote, A. Edward. "A Model of Communication Effectiveness." Journal of Communication 20, no. 1 (1970): 84-91. Gantt, Vernon W. "Attitude Change as a Function of Source Credibility and Levels of Involvement." DAI 31 (1970): 3074A (Ohio University). Garcia-Esteve, Joel. "Attitude Change as a Function of Extremity of Initial Position and Amount of Change Advocated." DAI 30 (1969): 2148-2149A (University of Florida). ‘TT . Garrett, Roger L. "Developmental Changes in the Effects of Communicator Position on Attitude Change." EA 29 (1969): 4102-4103A (Stanford University). Gerbnen George. "Communication and Social Environment." Scientific American 227, no. 3 (1972): 152-160. Gerlach, Vernon, and Ely, Donald. Teaching and Media: A Systematic Approach. Englewood Cliffs, N.J.: Prentice Hail, Ihc., 1971. Gillig, Paulette M., and Greenwald, Anthony 6. "Is It Time to Lay the Sleeper Effect to Rest?" Journal of Person- ality and Social Psychology 29, no. 1 (1974): 132-139. Ginter, James L. “An Experimental Investigation of Atti- tude Change and Choice of a New Brand." Journal of MarketingAResearch 11, no. 1 (1974): 30-40. Goldberg, Carlos I. "Attitude Change as a Function of Source Credibility, Authoritarianism, and Message Ambi- guity." EAL 31 (1970): 464-465A (City University of New York). 150 Goldsmith, Francis J., and Hochbaum, Godfrey M. "Changing People's Behavior Toward the Environment." Public Health Reports 90, no. 3 (1975): 231-234. Goulard, Cary J. "The Development of an Analysis Technique for Determining the Potential Effectiveness of Tele- vised Environmental Public Service Announcements to Cause Affective Change in Urban Adults.“ DAI 35 (1975): 58528 (State University of New YorETT Goldstein, E. Bruce. "The Perception of Multiple Images." Audio-Visual Communication Review 23, no. 1 (1975): 34-68. Gombrich, E.H. "The Visual Image." Scientific American 227, no. 3 (1972): 82-96. Greenberg, Bradley S., and Miller, Gerald R. "The Effects of Low-Credible Sources on Message Acceptance." Speech Monographs 33, no. 2 (1966): 127-136. Greenwald, Anthony G. "Behavior Change Following a Persua- sive Communication." Journal of Personality 33, no. 3 (1965): 370-391. Hale, Gordon A., Miller, Leon K. and Stevenson, Harold W. "Incidental Learning of Film Content: A Developmental Study." -Child Development 39, no. 1 (1968): 69-77. Hass, Jane W., Bagley, Gerrold S., and Rogers,Ronald W. "Coping With the Energy Crisis: Effects oleear Appeals Upon Attitudes Toward Energy Consumption." Journal of Applied Psychology 60, no. 6 (1975): 754-756. Hass, R. Glen. "Persuasion of Moderation? Two Experiments on Anticipatory Belief Change." Journal of Personalit and Social Psychology 31, no. 6 ( : - Hazen, Michael D., and Kiesler, Sara 8. "Communication Strategies Affected by Audience Opposition, Feedback and Persuasibility." Speech Monographs 42, no. 1 (1975): 56-68. Heberlein, Thomas A. "The Land Ethic Realized: Some Social Psychological Explanations for Changing Environmental Attitudes." Journal of Social Issues 28, no. 4 (1972): 79-87. Heller, Jack F., Pollak, Michael S., and Picek, James M. "The Interactive Effects of Intent and Threat on Boom- erang Attitude Change." Journal of Personality and Social Psychology 26, no. 1 (1973): 273-279. 151 Hempstead, John 0. "Media and the Learner: The Influence of Media-Message Components on Students' Recall and Attitudes Toward the Learning Experience." EAL 34 (1974): 6674A (University of Wisconsin). Hilyard, Delmer M. "One-Sided vs. Two-Sided Messages: An Experiment in Counterconditioning." EA 27 (1966): 1109-1110A. Himmelfarb, Samuel, and Arzai, Daniella. "Choice and Source Attractiveness in Exposure to Discrepant Mes- sages." Journal of Experimental Social Psychology 10, no. 6 (1974): 516-527. Holz, Robert F., and Rosnow, Ralph L. "Awareness of Expec- tation in Attitude Research." Psychological Reports 20, no. 2 (1967): 642. Horai, Joann, Naccari, Nicholas, and Fatoullah, Elliot. "The Effects of Expertise and Physical Attractiveness Upon Opinion Agreement and Liking." Sociometry 37, no. 4 (1974): 601-606. Hournshell, Paul B. and Liggett, Larry. "Assessing the Effectiveness of Environmental Education." The Journal of Environmental Education 5 (Winter 1973): 27-28. Hovland, Carl 1., Janis, Irving L., and Kelley, Harold H. Communication and Persuasion. New Haven: Yale University Press, 1953. Insko, Chester A., Turnbull, William, and Yandell, Ben. "Facilitative and Inhibiting Effects of Distraction on Attitude Change." Sociometpy 37, no. 4 (1974): 508-528. Isaacs, Morton. "Involvement as a Factor in Resistance to Persuasion." DAI 33 (1973): 3775A. Isaacson, Frederick W. "The Effects of Defensiveness and Involvement on the Acceptance of Fear Arousing Commun- ications." EAL 34 (1974): 6782-6783A (University of Denver). lsenhour, John P. "The Effects of Context and Order in Film Editing." Audio-Visual Communication Review 23, no. 2 (1975): 69-80. Jacobson, Harvey K. "The Credibility of Three Mass Media as Information Sources." EA 29 (1968): 3132A (Univer- sity of Wisconsin). 152 Jarvie, J.C. Movies and Society. New York: Basic Books, Inc., 1970. Johnson, Homer H., and Izzett, Richard R. "The Effects of Source Identification on Attitude Change as a Function of the Type of Communication." Journal of Social Psychology 86, no. 1‘(1972): 81-87. Kaplowitz, Stan A. "The Effects of Punishment, Power and Credibility on the Compliance of Threat Targets." DAI 32 (1972): 6537-6538A (University of Michigan). Kanter, Donald_L. "Communication Theory and Advertising Decisions." Journal of Advertising Research 10, no. 6 (1970): 3-8. Kay, Herbert. "Do We Really Know the Effects of Using 'Fear' Appeals?" Journal of Marketing 36, no. 2 (1972): 55-57. Keating, John P. "Persuasive Impact, Attitudes, and Images: The Effects of Communication Media and Audience Size on Attitudes Toward a Source and Toward His Advocated Position." EAL 33 (1973): 3981-3982 (Ohio State University). Kehoe, John W. "Application of Principle of Inconsistency to Strategies for Changing Attitudes Toward Culturally Diverse Groups." EAL 34 (1974): 4700-4701A (University of Tornoto). Kiesler, Sara 8., and Mathog, Roberta B. "Distraction Hypothesis in Attitude Change: Effects of Effective- ness." 3Psychological Reports 23, no. 3 (1968): 11 -11 3. Koenigil, Mark. Movies in Society. New York: Robert Speller and Sons Publishers, Inc., 1962. Kraus, S., El-Assal, Elaine, and DeFleur, M.L. "Fear- Threat Appeals in Mass Communication: An Apparent Con- tradiction." Speech Monographs 33, no. 1 (1966): 23-29. Kress, Stephen W. "A Study of the Modification of Chil- dren's Negative Attitudes Toward Animals." DAI 36 (1975): 2761A (Cornell University). Krugman, Herbert G. "Processes Underlying Exposure to Advertising." American Psychologist 23, no. 4 (1968): 245-253. 153 Lana, Robert E., and Rosnow, Ralph L. "Effects of Pretest Treatment Interval on Opinion Change." Psychological Reports 22, no. 3 (1968): 1035-1036. LaRossa, Ralph. "Interpreting Hierarchical Message Struc- ture." Journal of Communication 24, no. 2 (1974): 61-68. Laswell, Gordon. "Matrix of Persuasive Communication." In A Handbook of Social Psychology. Edited by Gardner Lindsey and Elliot Aronson, 2nd ed. Reading, Mass.: Addison-Wesley, 1969. League of Women Voters. Children's Impressions of American Indians: A Survey of SUBurban Kindergarten aha Fifth Grade Children. New Bfighton, Minn.: League of Women Voters, 1975. ERIC ED 134 499. . Young People's Concepts of Native Americans: A Survey of uEurban Ninth and Twelfth Grade Students. New Brighton, Minn.: League of Women Voters, 1976. ERIC ED 134 500. . Leopold, Aldo. A Sand County Almanac. New York: Ballantine Books, 1966. Leslie, Robert Franklin. In the Shadow of a Rainbow. New York: W.W. Norton and Company, Inc., 1974. Letchworth, George A. "Attitude Change Over Time as a Function of Ego Involvement, Communicator Credibility and Number of Exposures to the Communication. DA 29 (1969): 2802-2803A (University of Oklahoma). ‘— Leventhal, Howard. "Fear: For Your Health." Psychology Today 1, no. 5 (1967): 54-58. Lindsey, Gardner, and Aronson, Elliot, eds. A Handbook of Social PsychDIOQY. 2nd ed. Reading, Mass.: Addison- Wesley, 1969. Littlejohn, Stephen W. "An Experimental Study of Source Credibility and Communication Exposure." DAI 31 (1970): 30288 (University of Utah). London, Perry. Beginning Psychology. Homewood, Ill.‘ Dorsey Press, 19751 Lopez, Barry Holstun. Of Wolves and Men. New York: Charles Scribners' Sons, 1978. Lubalin, James S. "The Effects of Counterattitudinal Improv- ization on Attitudes: An Exploration of Mediating Proc- esses." DAI 31 (1970): 827A (University of Michigan). 154 Lucas, Pauline. "Status Consistency and Resistance to Change." EA 27 (1966): 831A (University of Michigan). Luchins, Abraham S., and Luchins, Edith H. "The Effects of Order of Presentation of Information and Explanatory Models." Journal of Social Psychology 80, no. 1 (1970): 63-70. Lyle, Jack. "Attitude Measurement in Communication Research." Journalism Quarterly 42, no. 4 (1965): 606-614. Madsen, Roy Paul. The Impact of Film. New York: Macmillan Publishinc Co., Inc., 1973. Maile, Carlton A. "Self-Esteem and Source Credibility as Determinants of Attitude Change." DAI 36 (1976): 5042A (University of Georgia). Marciniak, Kenneth A. "An Integrative Approach to Resist- ance to Attitude Change." DAI 34 (1973): 4418 (Loyola University, Chicago). Marcus, Alan M. "Audience Variables in Persuasion." DAI 30 (1970): 5515-5516A (University of Toronto). McCroskey, James C., and Wright, David W. "A Comparison of the Effects of Punishment-Oriented and Reward-Oriented Messages in Persuasive Communication." Journal of Communication 21, no. 1 (1971): 83-93. McCroskey, James C., and Combs, Walter H. "The Effects of the Use of Analogy on Attitude Change and Source Credibility." Journal of Communication 19, no. 4 (1969): 333-339. McCullough, Joseph L. "The Use of a Message of Net Counter- argumentation in Differentiating the Impact of Persua- sive Communications." EAL 32 (1972): 4274-42758 (Ohio State University). McEwen, William J. "Counter-Attitudinal Encoding Effects on Message Style and Performance." DAI 30 (1969): 2556A (Michigan State University). McEwen, William J.,and Greenberg, Bradley S. “The Effects of Message Intensity on Receiver Evaluations of Source, Message and Topic." Journal of Communication 20, no. 4 (1970): 340-350. McGuckin, Henry E., Jr. "The Persuasive Force of Similarity in Cognitive Style Between Advocate and Audience." Speech Monographs 23, no. 1 (1967): 145-151. 155 McGuire, William J. "The Nature of Attitudes and Attitude Change." In A Handbook of Social Psychology. 2nd ed. Edited by Gardner Lindsey and Elliot Aronson. Reading, Mass.: Addison-Wesley, 1969. McKillip, Jack, and Edwards, John D. "Source Character- istics and Attitude Change." Personality and Social Psychology Bulletin 1, no. 1 (1974): 135-137. McNally, Jeanne M. "A Study of Immediate and Delayed Retention as Related to Three Methods of Imparting Knowledge." DAI 30 (1969): 1872-1873A (Catholic University of‘Afierica). Mech, L. David. The Wolves of Isle Royale. Fauna Series, No. 7. Washington, D.C.: U.S. Government Printing Office, 1966. Messerschmidt, Ramona. "The Suggestibility of Boys and Girls Between the Ages of Six and Sixteen Years." IE3 Peda o ical Seminary and Journal of Genetic Psychology 43 (1933): 422-437. Miller, Gerald R., and Hewgill, Murray A. "Some Recent Research on Fear-Arousing Message Appeals." Speech Monographs 23, no. 4 (1966): 378-391. Mills, Judson. "Effect on Opinion Change of the Communica- tor's Liking for the Audience He Addressed." Psychonomic Science 25, no. 6 (1971): 335-337. Mills, Judson, and Harvey, John. "Opinion Change as a Func- tion of When Information About the Communicator Is Received and Whether He Is Attractive or Expert." Journal of Personality and Social Psychology 21, no. 1 11972): 52-55. Mintz, Stanley 1. "Attitude Change as a Function of Person- ality, Ego-Involvement and Transparency of the Influence Attempt." EAL 32 (1972): 5451-54528 (Texas Tech University). Mitchell, John 6. "Fear and Loathing in Wolf Country." Audubon 78, no. 3 (1976): 20-39. Moldstad, John A. "Selective Review of Research Studies Showing Media Effectiveness: A Primer for Media Direc- tors." Audio-Visual Communication Review 22, no. 4 (1974): 387-401. Mowat, Farley. Never Cry Wolf. New York: Dell Publishing Company, 1963. 156 Munson, Paul, and Kiesler, Charles A. "The Role of Attribu- tions by Others in the Acceptance of Persuasive Commun- ications." Journal of Personality 42, no. 3 (1974): 453-466. Nelson, Carnot E. "Anchoring to Accepted Values as a Tech- nique for Immunizing Beliefs Against Persuasion." Journal of Personality and Social Psychology 9, no. 4 ’(1968): 329-334. Norman, Ross. "When What Is Said Is Important: A Comparison of Expert and Attractive Sources." Journal of Ex eri- mental Social Psychology 12, no. 3 ( : - Ochs, Leonard A. "Primacy and Recency in Different Age Groups." DAI 34 (1974): 5172-51738 (State University of New YorE, Albany). Ostermeier, Terry H. "Effects of Type and Frequency of Reference Upon Perceived Source Credibility and Atti- tude Change." Speech Monographs 34, no. 2 (1967): 137-144. Paletz, David L., Koon, Judith, and Whitehead, Elizabeth, Hagens, Richard B. "Selective Exposure: The Potential Boomerang Effect." Journal of Communication 22, no. 1 (1972): 48-53. Pasquale, Thomas A. "Messages of Coercion: A Content Analysis of Power Themes." DAI 37 (1977): 5882-58838 (Syracuse University). Patterson, Cecil H. Insights About Persons: Psychological Foundations of Humanistic and Affective Education. Washington, D.C.: AssoCiation for Supervision and Cur- riculum Development, 1977. ERIC ED 133 887. Perry, Bonita L. "Differential Effort in Counterattitudinal Advocacy Effects on Attitude Change and Extinction." DAI 32 (1971): 1624-1625A. Pinson, Christian, and Roberto, Eduardo, L. "00 Attitude Changes Precede Behavior Change?" Journal of Adver- tising Research 13, no. 4 (1973): 33138. Pontius, Anneliese A. "New Frontiers of Ecological Ethics: A Balance Between Over- and Under-Control of Outer and Inner Factors." Experientia 31, no. 2 (1975): 263-264. Powell, Frederic A., and Miller, Gerald R. "Social Approval and Disapproval Cues in Anxiety Arousing Communica- tion." Speech Monographs 24, no. 2 (1967): 152-159. 157 Pyszka, Ronald H. "Cognitive Discrepancy and the Avoidance of Threatening Information." EA 28 (1968): 3266A (University of Wisconsin). Rand, Melvin A. "An Empirical Comparison of Sherif's Social Judgment Approach and Festinger's Dissonance Theory at Their Points of Contrast: Ego Involvement and Discrep- ancy of Communication." EA 28 (1967): 1234-12358 (University of Oklahoma). Reinsch, N. Lamar. "An Investigation of the Effects of the Metaphor and Simile in Persuasive Discourse.“ Speech Monographs 38, no. 2 (1971): 142-145. Reymert, Martin, and Kohn, Harold A. "An Objective Investi- gation of Suggestibility." Character and Personality 9 (1941): 44. Rhine, Ramon J., and Polowniak, William A. "Attitude Change, Commitment and Ego Involvement." Journal of Personal- ity 9 (1941): 44. Roberts, John C. "Group Engendered Attitude Change." DAI 35 (1975): 56258 (Purdue University). Rosen, Theodore J. "The Role of Fear and Danger in Persua- sion Following a Fear Communication." DAI 34 (1973): 416A (University of Wisconsin). Rosnow, Ralph L. "One-Sided Versus Two-Sided Communication Under Indirect Awareness of Persuasive Intent." Public Opinion Quarterly 32, no. 1 (1968): 95-101. . "Whatever Happened to the 'Law of Primacy?'" Journal of Communication 16, no. 1 (1966): 10-31. Rosnow, Ralph L., and Robinson, Edward J. Experiments in Persuasion. New York: Academic Press, 1967. Rossomando, Nina P. "Attitudinal Effects of Source Charac- teristics, Type of Commitment, and Choice to Listen to Dissonant Information." EA; 32 (1971): 3425A (City University of New York). Rudin, Lawrence A., and Hood, William. "The Point of Fal- lacy and Latitude Dimensions of Attitude Change." Journal of Social Psychology 87, no. 2 (1972): 235-241. Rupnow, David E. "A Study of the Role of Ego-Involvement in Predicting the Attitude Change of an Individual Subject as a Result of Viewing One or More Persuasive Filmscwla Controversial Issue." DAI 36 (1976): 5186A (University of Wisconsin, Madison)."—' 158 Schillaci, Anthony, and Culkin, John M., eds. Films Deliver: Teaching Creativity With Film. New York: Citation Press, 1970. Schramm, Wilbur. Men, Messages and Media: A Look at Human Communications. *New York: Harper & Row, 1973. Schramm, Wilbur, and Roberts, Donalds, eds. The Process and Effects of Mass Communication. Urbana, 111.: University of IIIinois Press, 1974. Schulman, Gary 1. "The Popularity of Viewpoints and Resistance to Attitude Change." Journalism Quarterly 45, no. 1 (1968): 86-90. Schweitzer, Don A. "Style of Presentation as a Credibility Variable." EA 28 (1968): 4273-4274A (University of Nevada). Scileppi, John A. "The Effects of Source Credibility and Ego Involvement on Attitude Change Toward a Descrepant Communication Under Facilitative and Non-Facilitative Cognitive Response Expression Conditions: An Elabora- tion of the Evaluative Set Theory Approach." EAL 34 (1973): 4458 (Loyola University, Chicago). Scollon, Robert W., Jr. "A Study of Some Communicator Vari- ables Related to Attitude Restructuring Through Motion Picture Films." EA 17 (1957): 400 (Pennsylvania State University). Sears, David 0., and Abeles, Ronald P. "Attitudes and Opgnions." Annual Review of Psychology 20 (1969): 25 -288. Shosteck, Herschel. “The Structural Dimensions of Tele- vision Editorial Effectiveness." Journalism Quarterly 52, no. 1 (1975): 37-43. Simmons, Tim. "Non-Violence Advocated." Michigan State News, 2 May 1979, p. 1. Skolnick, Paul, and Heslin, Richard. "Approval Dependence and Reactions to Bad Arguments and Low Credibility Sources." Journal of Experimental Research in Person- ality 5, no. 3 (1971): 199-207. Smith, M. Brewster. "Environmental Degradation and Human Behavior: Can Psychology Help?" Representative Research in Social Psychology 4, no. 1111973): 2271234. 159 Souers, Charles V. "Incidental Treatment Effects on Student Attitudes Toward a Current Social Problem." DAI 33 (1973): 4955-4956A. Spillman, Bonnie M. "Cognitive Dissonance, Self-Esteem, and the Process of Self Persuation." DAI 35 (1974): 1776-1777A (University of Utah). ‘ Stevenson, Harold W., and Siegel, Alexander. "Effects of Instructions and Age on Retention of Filmed Content." Journal of Educational Psychology 60, no. 1 (1969): 71-74. Stewart, Roy T. "An Experimental Investigation of the Relationship Between Perceived New Information and Delayed Attitude Change." EAL 34 (1974): 7916A (University of Illinois). Stone, Vernon A. "Personality Factors in Immunizing Against Source or Content of Persuasive Messages." EA 28 (1967): 1129A (University of Wisconsin). Stuteville, John R. "Psychic Defenses Against High Fear Appeals: A Key Marketing Variable." Journal of Marketing 34, no. 2 (1970): 39-45. Tate, Eugene D. “A Comparison of the Relative Immunizing Effect of Counterattitudinal Advocacy With the Passive Reception of a Persuasive Message." EAL 32 (1971): 538A (Michigan State University). Telaak, Kathleen A. "The Effect of Issue Involvement and Discrepancy on Latitude with Width as an Individual Variable in Attitude Change." DAI 32 (1972): 5501-55028 (University of MassaEfiUsetts). Touhey, Catherine E. "Prior Information, Credibility and Attitude Change." Journal of Social Psychology 95, no. 2 (1975): 287-288. Truzzi, Marcello. "Explorations in Cognitive Balance Theories: Experimental Assessments of Conflicting Theoretical Predictions in the Cognitive Consistency Theories of F. Heider, T.M. Newcomb, and C.E. Osgood and P.H. Tannenbaum." .EAE 31 (1970): 3063A. . "The Problem of Relevance Between Orientations for Cognitive Dissonance Theory." Journal for the Theory of Social Behavior 3, no. 2 (1973): 239-247. Tubbs, Stewart L. "Explicit Versus Implicit Conclusions and Audience Commitment." Speech Monographs 35, no. 1 (1968): 14-19. 16D Tuppen, Christopher J. "Perceptions of Source Credibility and Change in Attitudes." DAI 32 (1971): 1075-1076A (University of California, Berkeley). - Utz, Walter J., Jr. "The Comparative Effect of Color and Black and White Film Clips Upon Rated Perception of Reality." EAl 29 (1969): 2664A (University of Illinois). Vickers, Geoffrey. "Levels of Human Communication: Analysis of Three Functions of Communication." Communication 1, no. 1 (1974): 7-24. Vohs, John L., and Garrett, Roger L. "Resistance to Persua- sion: An Integrative Framework." Public Opinion Quarterly 32, no. 3 (1968): 445-452. Walker, Robert V., and Chalmers, Douglas K. "The Use of Technical and Non-Technical Language in Communication." Journal of Psychology 78, no. 1 (1971): 65-71. Wall, Victor 0. "Balance Theory, Ego-Involvement and Atti- tude Change." DAI 32 (1972): 4156-4157A (University of Colorado). Wall, Victor 0., and Boyd, John A. "Channel Variation and Attitude Change." Journal of Communication 21, no. 4 (1971): 363-367. Wallace, William F. "Primacy and Recency Warnings and Order Effects in Persuasive Communications." DAI 31 (1971): 43758 (University of Alabama). Ward, Charles 0., and McGinnies, Elliott. "Persuasive Effects of Early and Late Mention of Credible and Non- credible Sources." Journal of Psychology 86, no. 1 (1974): 17-23. Warshofsky, Michael W. "The Effect of a Subject-Communica- tor Attitudinal Similarity-Dissimilarity and Communica- tion Discrepancy on Attitude Change and Demand Characteristics." DAI 32 (1971): 11998 (University of Nebraska). -_' Weber, Stephen J. "Opinion Change as a Function of the Associative Learning of Content and Source Factors." DAI 33 (1972): 27988 (Northwestern University). . "Source Primacy-Recency Effects and the Sleeper Effect." Proceedings of the Annual Convention of the American PsychoIogical Association 6111971): 385-386? 161 Weintraub, Roy L. "The Motion Variable in Film Presenta- tion." DAI 34 (1973): 562-562A (University of Southern California). . Weiss, Paul. Cinematics. Carbondale, 111.: Southern Illinois Uhiversity Press, 1975. Weiss, Walter. "Communicator Effectiveness in Relation to the Strength of the Communication." Psychological Reports 20, no. 3 (1967): 1037-1038. Wells, Caryll S. "A Comparison of Three Models of Attitude Change." EA 29 (1967): 1037-1038. Wenburg, John R. "The Relationships Among Audience Adapta- tion, Source Credibility and Types of Message Cues." DAI 30 (1970): 5555-5556 (Michigan State University). Weston, John R. "Argumentative Message Structure and Prior Familiarity as Predictors of Source Credibility and Attitude Change." EA 28 (1968): 3772A (Michigan State University). Wheeless, Lawrence R. "The Effects on Attitude Credibility and Homophily on Selective Exposure to Information.9 Speech Monographs 41, no. 4 (1974): 329-338. Wiebe, Gerhart 0. "Two Psychological Factors in Media Audience Behavior." Public Opinion Quarterly 33, no. 4 (1970): 523-536. Williams, Robert C. "Film Shots and Expressed Interest Levels." Speech Monographs 35, no. 2 (1968): 166-169. Williams, Tess M. "A Procedure for Utilizing Rhetorical Theory and the Findings of Behavioral Science and Com- munication Research in the Design of 'High-Yield' Per- suasive Messages." EAl 32 (1972): 7120A (Michigan State University). Windley, Paul G. "Environmental Dispositions of Older People." DAI 34 (1973): 17378 (University of Michigan). Woods, Donald J. "Repression-Sensitization, Attitude Towards Emotionality and Response to a Threatening and Non-Threatening Film." DAI 32 (1972): 5464-54658 (Northwestern UniversityTT_ Worchel, Stephen. "The Effects of Films on the Importance of Behavioral Freedom." Journal of Personality 40, no. 3 (1972): 417-435. 162 Wright, Peter, and Barbour, Frederic. "The Relevance of Decision Process Models in Structuring Persuasive Mes- sages." Communication Research 2, no. 3 (1975): 246-259. Yontef, Gary M. "Communication Displacement, Communicator - Evaluation and Attitude Change as a Function of Inwflve- ment, Discrepancy and Order of Presentation." EAl 31 (1971): 5649B. Young, Stanley P., and Goldman, Edward A. The Wolves of North America. 2 vols. New York: Dover Publications, Inc., 1944. Zeigler, Sherilyn K. "Attention Factors in Televised Mes- sages: Effects on Looking Behavior and Recall." DAI 30 (1969): 2625-2626A (Michigan State University). Zimbardo, Philip, and Ebbesen, Ebbe 8. Influencing Atti- tudes and Changing Behaviors. Reading, Mass.: Addison- Wesley Publishing Company, 1970. Zimmerman, Peggy. "Retention in Threatening and Non- Threatening Stories by Kindergarten Children." Gradu- ate Research in Education and Related Disciplines 4, no. 1 (1968): 81-91? APPENDICES APPENDIX A APPENDIX A PILOT STUDY RESULTS Cooperation Indicators Two questions were included within the survey to estimate the sincerity of the children responding to the test items. The results indicate that well over 85% of all grades involved were pleased to have participated in the survey. This finding suggests that the students may have made a sincere attempt to register their true feelings regarding the test items. Knowledge of Animals It is of interest to note that approximately one-third of the third grade, one-half of the fourth grade, and three- fourths of the fifth-grade children knew that humans were members of the animal family. This appears to demonstrate a knowledge element that is generally acquired by fifth grade but is less well known prior to this time. Over 50% of all respondents believe that people can live in harmony near wild animals, but a substantially lower percentage thought that the inverse was true. Also, over 50% of all respondents believe that hunting wild ani- mals should not be allowed. When asked about animal intel- ligence, 75% of all students surveyed indicated that they believed wild animals were not lacking in intelligence. 163 164 Ninety percent of the children believe it is wrong for man to deny the use of land areas to wild animals. Attitude Toward Animals Substantially less than 50% of all children tested felt that man was the most important animal species on earth. The greatest variance occurred at the fourth-grade level, with only 16% in agreement with the statement. Several questions were included in the survey to determine if children perceived wild animals as being inher- ently "good or bad." Approximately 75% of all respondents felt that in nature there are good and bad animals. Deer were the most popular wild animals, perceived as being "good" by 100% of the third graders and well over 80% of the fourth and fifth graders. Bear were generally not thought of as being good animals. Only 28% of the third graders, 20% of the fourth graders, and 18% of the fifth graders indicated that the bear is a "good" animal. Of the two domestic species utilized within the survey, horses were thought to be "good" animals by 72% of the third graders, 64% of the fourth graders, and 81% of the fifth graders. The cow was also considered to be a "good" animal but to a lesser degree than the horse. Slightly over 50% of the third and fourth-grade children, and 70% of the fifth graders tested, felt the cow was a "good" animal. When the good/bad question is applied to the wolf, the children reacted in an expected manner. Only 8% of the third graders tested felt that the wolf was a "good" 165 animal. Approximately 20% of the fourth and fifth graders also felt the wolf was "good." Fifty-six percent of the third graders felt that the wolf was not a "good" animal. It is interesting to note that 56% of the fourth graders and 40% of the fifth graders indicated that they really did not know the answer to this question. Fifty percent of the children tested indicated that they believed dogs and wolves are physically quite dissimi- lar in appearance. Also, 50% of the children believe that the wolf is not a natural enemy of man. However, 50% of the third graders believed that it was. Over 80% of the sample at each grade level felt that wolves should not be eliminated by the effects of man's occupancy of their natural habitat. . Sixty-five percent of the third and fifth graders tested did not think a wolf would make a very good pet. About 40% of the fourth graders felt the same way, with another 40% unsure of the answer to this question. Two- thirds of the children indicated that they were not afraid of wolves. Concerning a question on how people and wolves get along, two-thirds of the third graders indicated that they felt they did not get along. When asked about "potential trouble" wolves may cause in a forest, the responses again prove interesting. While 50% of the third graders believe that wolves are beneficial to the forest, nearly 50% of the fourth graders disagree. This trend reverses itself once again in the fifth grade, 166 with 50% of the respondents indicating wolves were beneficial. The third grade appeared polarized on whether man is the enemy of the wolf. Nearly 50% either agreed or dis- agreed with the question. Very few children indicated a neutral position. Approximately fifty percent of the fourth graders were unsure of this question, while 40% of the fifth graders agreed. Nearly 50% of the third and fourth graders and 60% of the fifth graders felt that the wolf's ability to adapt to differing habitats was rather limited. Approximately 50% of the third graders believe that wolves only live deep within forest regions. Of the fourth and fifth graders, nearly 50% indicated that they were unsure about where wolves live. Fifty-six percent of the third graders knew' that wolves today are not as plentiful as in the past. However, 60% of the fourth graders and 40% of the fifth graders indicated that they did not know the answer to this question. Operating upon the assumption that most children are familiar with the story "Little Red Riding Hood," a ques- tion was included in the survey to determine the story's effect upon the children. The third graders appeared to have the strongest opinion, with 36% of them indicating that the story had the correct idea concerning wolves. However, 48% of the third graders also felt the story was inaccurate in depicting the wolf. A general shift in 167 attitude appears in the fourth graders. When asked the same question, 65% of the fourth graders indicated that they disagreed with the portrayed image of the wolf. Fifth graders demonstrated considerable uncertainty on this ques- tion, with 58% indicating that the story did not accurately depict the wolf. However, 25% of the fifth graders tested were in agreement with the story's wolf portrayal. When asked if retribution should be employed against wolves which kill deer, 56% of the third graders thought not. The rest of the class was about evenly divided on this question, with 24% indicating that wolves should be killed for killing deer. The remainder of the class neither agreed nor disagreed with the question. None of the fourth graders felt that retribution should be employed against deer-preying wolves. Interestingly, 84% of the fourth grade children felt that it was not fit- ting to kill wolves who prey upon deer. The fifth grade sample indicated that 70% would not seek retribution for deer-killing wolves. Twenty percent were unsure of their response to this question, while the remaining 10% would have the offending wolf killed for its predaciousness upon deer. When asked if perhaps God made a mistake in creating the wolf, 75% of all children tested felt that He had not. Only 3% of the combined classes felt that God had perhaps erred in His creation of the wolf. The creation question was reworded utilizing the term Nature in place of God. 16B Eighty percent of the third and fourth graders and 76% of the fifth grade believed Nature had not erred. Two percent of the total sample felt that Nature had made a mistake in creating the wolf. Seventy-five percent of the third graders felt that wolves should not be hunted for sport. However 16% felt it was acceptable to do so. The fourth grade responded to this question with 90% indicating that hunting wolves was not desirable. None of the fourth-grade children indicated that sport hunting of wolves was an acceptable practice. The fifth graders were also opposed to sport hunting of wolves, with 85% so indicating. Only 4% 0f the fifth graders thought it an acceptable practice. When asked whether it was proper for mankind to uti- lize the planet for whatever purposes desired, the children responded in an interesting manner. Twenty-five percent of the third graders, 8% of the fourth graders, and only 1% of the fifth graders felt that man should utilize the earth for his exclusive gain. Approximately 60% of the third graders and 80% of the fourth and fifth graders disagreed with this position. Nearly 90% of all respondents indicated that man should protect wild animals. Attitude Towards Message Source/Authority Figures Several questions were included in the survey to sam- ple the respondents' attitudes towards a message source and authority figures in general. When asked about believing 169 everything that they read in books, 76% of the third graders, 96% of the fourth graders, and 85% of the fifth graders indicated that they felt they should not believe everything that they read. Between 85% and 90% of all children also felt that TV and movies did not always deliver accurate information. When the same question was asked with regard to the believability of adults, the children responded in quite a different manner. Nearly 50% of the third graders felt that they should believe everything adults say. Only 24% of those third graders tested felt that they should ques- tion an adult's accuracy. This figure increases to 52% for the fourth grade, and peaks at 75% in the fifth-grade 'sample. When asked if they should believe everything their teachers say, 70% of the third graders agreed. An attitude shift appears to begin to form in the fourth grade with over 50% of the respondents unsure as to whether they believe everything teachers say. This attitude shift appears to be fully established by fifth grade, with well over 50% of the children indicating that they do not believe everything their teachers say. Between 85% and 97% of all respondents indicated that they had seen pictures of wolves. Only 16% of the third graders, 30% of the fourth graders, and approximately 50% of the fifth graders indicated that they had never seen a real wolf. Conversely, 36% of the third, 20% of the 170 fourth grade, and 15% of the fifth grade claimed to have encountered wolves while in a woods or forest. APPENDIX B Instructions: Example: Practice: APPENDIX B PILOT STUDY OPINION SURVEY Room Number Teacher School Grade Male Female OPINION SURVEY Please read each question and circle the number that shows how you feel about it. Circle only one number for each question. Answer all questions. A. Michigan State University is in East Lansing, Michigan 1. -------------- 2. ----------- 3. (agree) (neither) (disagree) If you a ree with it, circle #1. If you aisa ree with it, circle #3. If you neither agree or disagree with it, circle #2. 8. There are times when school is fun. 1. -------------- 2. ----------- 3. (Agree) (neither) (disagree) Please answer the Example 8. question at this time. 171 10. 11. 12. 13. 172 People are members of the animal family. Dogs and wolves are so much alike its difficult to tell them apart. 1. ----------- 2. ---------- 3. Horses are considered to be good animals. 1. ----------- 2.- --------- 3. I probably should believe everything I read in books. 1. ----------- 2.- --------- 3. Wild animals can't live in peace near people. 1.-- --------- 2. ---------- 3. The wolf is the natural enemy of man. 1 ------------ 2. ---------- 3. To make room for more people on earth, wolves will have to be killed. 1. ----------- 2. ---------- 3. Deer are considered to be bad animals. I probably should believe everything I see on TV and in movies. 1. ----------- 2. ---------- 3. People should be allowed to hunt animals. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 173 If a wolf kills a deer, the wolf should be killed. want. People and wolves just do not get along. 1 ------------ 2. ---------- 3. Wolves cause more trouble than good in the forest. 1. ----------- 2. ---------- 3. Man is the enemy of the wolf. 1. ----------- 2. ---------- 3. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 174 In Nature, there are good animals and bad animals. Before man clears a forest for his use, he had better kill the wolves that live there. It is right that we take the land away from the wild animals for our use. 1. ----------- 2. ---------- 3. There are more wolves alive today than ever before. 1. ----------- 2. ---------- 3. Wolves live only in the frozen north. 1. ----------- 2. ---------- 3. The story about Little Red Riding Hood had the right idea about wolves. Maybe Nature made a mistake when it made the wolf. 1. ----------- 2. ---------- 3. I think wolves should be hunted for sport. 1. ----------- 2. ---------- 3. I probably should believe everything my teachers say. 175 40. I have seen wolves in a 200. 41. I have seen wild wolves in the woods or forest. 1. ----------- 2. ---------- 3. 42. I have never seen a real wolf. 1. ----------- 2. ---------- 3. 43. I'm glad I had a chance to give my true feelings about these questions. APPENDIX C APPENDIX C ANSWERS TO PILOT SURVEY , Grade 3 Grade 4 Grade 5 § Grade Choice Choice Choice 1:: 5% E 3 4 5 1 2 3 1 2 3 1 2 3 '5 In 3 % in % response % response % response 3 35 g agreement each question each question each question a E '5 with judges 1 1 K 36 48 77 36 24 4O 48 8 44 77 1O 11 2 1 K 36 16 44 36 32 32 16 44 40 44 29 24 3 2 A 20 28 16 52 20 8 52 28 16 71 16 11 4 3 K-A 48 68 69 36 16 48 16 16 68 23 6 69 5 3 K 56 56 53 28 16 56 32 12 56 27 13 53 6 2' A - 24 20 16 72 24 4 64 20 16 81 16 1 7 3 A 76 96 85 8 16 76 0 4 96 0 13 85 ' 8 3 K-A 32 48 47 36 32 32 28 24 48 31 21 47 9 3 K-A 40 48 52 44 16 - 4O 20 21 48 19 27 52 10‘ 3 K-A 84 88 81 4 12 84 4 8 88 4 13 81 11 2 A O B 14 0 O 100 4 8 88 1 11 85 12 3 A 84 92 92 8 8 84 4 4 92 O 6 92 13 1 K-A 20 8 21 20 24 56 B 40 52 21 26 52 14 3 A 56 84 71 13 20 56 0 16 B4 10 19 71 15 3 A 40 16 4O 28 32 4O 20 64 16 9 42 4O 16 3 K-A 88 84 77 4 8 88 4 12 84 4 16 77 17 3 A 56 80 76 24 20 56 8 12 80 1 23 76 18 3 A 88 92 82 8 4 88 4 4 92 6 11 82 1 9 3 A 56 24 4O 8' 36 56 20 56 24 18 4O 4O 20 3 A 24 52 74 44 32 24 16 28 52 1 1 13 74 21 3 A 76 72 74 O 20 76 8 20 72 3 1B 74 22 3 K 64 44 66 16 16 64 16 40 44 1 1 23 66 23 3 K-A 60 72 66 16 24 6O 8 20 72 1 1 21 66 24 1 K-A 64 32 35 64 20 16 32 26 32 35 42 23 25 3 A 36 44 48 44 20 36 12 44 44 1 6 34 48 176 177 ANSWERS TO PILOT SURVEY continued Grade 3 Grade 4 Grade 5 at €53 Grade Choice Choice Choice 5 5 E 3 4 5 1 2 3 1 2 3 1 2 3 E 73 g % in % response % response % response a, c» E agreement each question each question each question a g a with judges 26 1 K 40 24 39 4O 4 56 24 40 36 39 29 32 27 3 A 8 16 18 72 20 8 76 8 16 74 6 18 B 1 Coop 88 100 90 88 12 O 100 0 0 90 O 4 28 1 K 40 24 18 40 16 44 24 28 48 18 13 58 29 3 A 44 72 65 44 12 44 12 16 72 15 19 65 30 3 K 24 28 17 44 32 24 16 44 28 37 44 18 31 2 K-A 0 4 3 8 0 92 4 4 92 8 3 ‘ 87 32 3 K 56 24 35 16 28 56 12 6O 24 23 40 35 33 3 K 80 92 72 4 16 80 O 4 92 8 18 72 34 3 A 48 64 56 36 16 48 20 16 64 26 15 56 35 3 K 44 24 29 16 - 4O 44 24 52 24 18 52 29 36 3 A 80 BO 76 4- 16 80 5 16 80 2 18 76 37 3 A 76 88 85 16 8 76 O 12 88 4 8 85 38 3 A 20 24 54 68 12 20 24 52 24 24 19 54 39 84 0 12 92 8 O 97 O 3 4O 64 8 16 84 0 16 58 8 32 41 3 56 60 71 36 8 56 20 20 6O 15 11 71 42 16 8 76 32 0 68 44 6 50 43 Coop 84 96 85 84 12 3 96 4 0 85 1O 2 APPENDIX D 10. 11. 12. SO SO SO SO SA Cog. Cog. Cog. Cog. Aff. Aff. Cog. Cog. Aff. Cog. Aff. Aff. APPENDIX D PRETEST - TEST #1 Unlike dogs, wolves usually don't get fleas or other bugs. In a natural forest, every animal serves a purpose. A wolf maps its territory with scent marks. A forest has no limit to the number of ani- mals it can support. When a wolf kills a deer in the forest, a hunter should kill that wolf. Natural wild areas are not needed for human survival. Scientists believe that wolves give gifts to each other. The white man's arrival in this country had no effect on the natural forest. Things that happen in a natural forest are: right or wrong and good or bad. Wolves are not very smart animals. I believe that wolves and other predators got what they deserved from people. The American Indian understood much about the land that people living today may never know. 178 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. SO SO SO SO SO Aff. Aff. Aff. Cog. Aff. Cog. Aff. Cog. Cog. Cog. Cog. Cog. Cog. Cog. 179 Without humans, there could be no natural forest or wild animals. Wilderness and Indians are things of the past and are no longer important. People no longer need wilderness areas. Wolves hunt throughout all areas of a natural forest. Animals that kill and eat others deserve no place in a natural forest. Wolves wander throughout all areas of a natural forest. In a forest, there are good animals and bad animals. Wolves use their body waste as messages for other wolves. Wolves were killed because they were thought to be wicked beasts. People and wolves often get the same kinds of sickness. In a natural forest there are animals called predators which kill and eat others. Although wolves kill, they are important in a natural forest. Wolves usually kill only the sick, old, and young. Wolves help control the number of animals within a natural forest. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 49. SA SO SO Cog. Cog. Cog. Cog. Cog. Aff. Cog. Cog. Cog. Aff. Aff. Cog. Cog. Cog. 180 When wolves live in a natural forest, the forest is doing well. Wolves learn most about the forest by seeing. Humans have a language, but wolves have none. Scent marks are the cause of different wolf packs fighting. Wolves enjoy the company of other wolves within their pack. I agree with the American Indian's beliefs about wolves and other animals. More than one pair of wolves within a pack each year have pups. Wolves will always kill deer when they have the chance. Most wolf pups born in a natural forest live to become adults. For a forest, the wolf is trouble. The story "Little Red Riding Hood" had the right idea about wolves. Many creatures in the forest hunt the wolf. Wolves from different packs often have friendly meetings within a forest. The white settler was willing to share the land with the wolf. APPENDIX E 10. 11. 12. 13. SO SO Cog. Cog. Aff. Cog. Cog. Aff. Cog. Cog. Aff. Cog. Cog. . Cog. Cog. APPENDIX E POSTTEST - TEST #2 A wolf maps its territory with scent marks. Wolves learn most about the forest by seeing. I agree with the American Indian's beliefs about wolves and other animals. The white man's arrival in this country had no effect on the natural forest. Wolves enjoy the company of other wolves within their pack. Wilderness and Indians are things of the past and are no longer important. Wolves usually kill only the sick, old, and _young. Scent marks cause different wolf packs to fight. When a wolf kills a deer in the forest, a hunter should kill that wolf. Wolves wander throughout all areas of a natural forest. Humans have a language but wolves have none. Wolves from different packs often have friendly meetings within a forest. Wolves use their body wastes as messages for other wolves. 181 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. SO SO SO SO SO Aff. Aff. Cog. Aff. Aff . Cog. Cog. Cog. Cog. Cog. Aff. Cog. Cog. 182 The story “Little Red Riding Hood" had the right idea about wolves. Natural wild areas are not needed for human survival. The white settler was willing to share the land with the wolf. I believe that wolves and other predators got what they deserved from people. Animals that kill and eat others deserve no place in a natural forest. Scientists believe that wolves give gifts to each other. A forest has no limit to the number of animals it can support. In a natural forest, there are animals called predators which kill and eat others. Unlike dogs, wolves usually don't get fleas or other bugs. I Wolves help control the number of animals within a natural forest. Without humans, there could be no natural forest or wild animals. More than one pair of wolves within a pack each year have pups. Wolves will always kill deer when they have the chance. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. SA SO SO SO SO Aff. Cog. Cog. Aff. Aff. Aff. Aff. Cog. Cog. Cog. Cog. Cog. Cog. Cog. 183 The American Indian understood much about the land that people living today may never know. Wolves are not very smart animals. Many creatures in the forest hunt the wolf. People no longer need wilderness areas. In a forest there are good animals and bad animals. For a forest, the wolf is trouble. Things that happen in a natural forest are right or wrong and good or bad. Although wolves kill, they are important in a natural forest. Most wolf pups born in a natural forest live to become adults. When wolves live in a natural forest, the forest is doing well. Humans and wolves often get the same kinds of sickness. Wolves were killed because people believe they are wicked beasts. Wolves hunt throughout all areas of a natural forest. In a natural forest, every animal serves a purpose. 184 Survey 41. A. I learned from the film. 42. A. I now feel differently about wolves, partly because of this film. 43. A. y I believe the story the Indian told. 44. A. I would like to see more of these film stories from Source Lake. APPENDIX F 10. 11. 12. 13. 14. SO SO Aff. Aff. . Cog. Cog. . Cog. Cog. Cog. Cog. Aff. Cog. Cog. Aff. Cog. Cog. APPENDIX F DELAYED POSTTEST - TEST #3 For a forest, the wolf is trouble. Things that happen in a natural forest are either right or wrong and good or bad. Wolves will always kill deer when they have the chance. Most wolf pups born in a natural forest live to become adults. Wolves learn most about the forest by seeing. Wolves help control the number of animals within a natural forest. Wolves enjoy the company of other wolves within their pack. Humans and wolves often get the same kinds of sickness. People no longer need wilderness areas. Many creatures in the forest hunt the wolf. A wolf maps its territory with scent marks. The American Indian understood much about the land that people living today may never know. Wolves were killed because people believe they are wicked beasts. Wolves are not very smart animals. 185 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. SO SO SA SD Cog. Cog. Cog. Cog. Cog. Aff. Cog. Cog. Cog. Aff. Cog. Aff. Aff. Cog. 186 Wolves from different packs often have friendly meetings within a forest.- Scent marks cause different wolf packs to fight. The white settler was willing to share the land with the wolf. Wolves use their body waste as messages for other wolves. Unlike dogs, wolves usually don't get fleas or other bugs. Natural wild areas are not needed for human survival. Wolves hunt throughout all areas of a natural forest. In a natural forest, there are animals called predators which kill and eat others. When wolves live in a natural forest, the forest is doing well. The story "Little Red Riding Hood" had the right idea about wolves. Humans have a language but wolves have none. I agree with the American Indian's beliefs about wolves and other animals. Wilderness and Indians are things of the past and are no longer important. Wolves usually kill only the sick, old, and young. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. SO SO SO SO SO Aff. Cog. Aff. Aff. Cog. Cog. Aff. Cog. Cog. Aff. Cog. Cog. 187 When a wolf kills a deer in the forest, a hunter should kill that wolf. Wolves wander throughout all areas of a natural forest. I believe that wolves and other predators got what they deserved from people. Animals that kill and eat others deserve no place in a natural forest. Scientists believe that wolves give gifts to each other. A forest has no limit to the number of animals it can support. Without humans, there could be no natural forest or wild animals. More than one pair of wolves within a pack each year have pups. In a natural forest, every animal serves a purpose. In a forest, there are good animals and bad animals. Although wolves kill, they are important in a natural forest. The white man's arrival in this country had no effect on the natural forest. Survey 41. A. 42. A. 43. A. 44. A. 188 I learned from the film. I now feel differently about wolves partly because of this film. I believe the story the Indian told. I would like to see more of these film stories from Source Lake. APPENDIX G APPENDIX G LETTER TO PARENTS OF CHILDREN PARTICIPATING IN STUDY Dear Parent, Your child's class has been invited to participate in an Ecology Atti- tude Study to be conducted by Mr. Peter J. Brobeil, a doctoral candi- date at Michigan State University. The investigation will require a total of two hours of school time, consisting of four one-half hour segments spread over a thirty day period. Children participating in the study will be administered a brief knowledge/attitude survey regarding the ecological issue under investi- gation. One week after the initial survey, the participants will view a thirty-minute documentary film and retake the knowledge/attitude sur- vey. Two weeks after viewing the documentary film, the children will again be administered the knowledge/attitude survey, and this will then conclude their involvement in the study. It is anticipated that participating students will not only enjoy the experience but will also gain valuable factual information regarding an endangered wildlife species. This research project has been approved by the Michigan State Univer- sity Committee on Research Involving Human Subjects (UCRIHS) and has the support of the administration of the East Lansing Public Schools. Results of the study will be conveyed to the school administration and may contribute to ongoing curriculum improvement. If you have any questions regarding this research or do not wish your child to participate, you are asked to contact the principal at Pinecrest School (337-2042). In addition, I will be available at Pinecrest during the morning hours, and would be happy to answer any questions you might have. I appreciate your cooperation in this endeavor. Sincerely, Peter J. Brobeil 189 APPENDIX H APPENDIX H SCENES AND NARRATION OF EXPERIMENTAL FILM: Scenes Boat Lake shore Moose/meadow Cabin Pete Pete/map Canoe Camp Summer cabin Winter cabin Cross lake Cross lake Bear Cross lake Grey jay Winter CIRCLE SONG By PETER BROBEIL Narration I remember---it was during the warm months ---just a few years back. I'd been staying in at the MacCoy cabin--- up on Source Lake. Been living out there alone---for several months. All that wilderness---and just one man--- it's a neat feeling. I like the warm months best when I'm out. That's the easiest time. The winds blow warm---and the long days make life pretty nice. But, I've also lived in this great forest during the agony of winter. Here where the cold hangs on the land like a great lead weight. The woods lie silent. Time suspended. The wandering bear of summer---now lies sleeping---deep within its den. But the forest is alive---it adjusts---it goes on. Most get by---most of the time. Whiskey Jack---knows how to get by. He'll take a handout---once he decides you're not going to eat him. I think that's what winter is all about out here---just getting by. For man it's always hard. 190 Summer Stream Chick/nest Chick/swim Hawk to nest Pete/Reed bird Pete Egg hatch Bears Wolves Pete Bear Chick/adult Sunset Cliffs Native Native Native Flying crane Native Forest animals 191 But as sure as the fall carried the breath of winter---the winter breathes life into spring. And the cycle continues---as it has for a thousand and more Years. During my stay at Source Lake---I grew in understanding and acceptance. Much of my time was spent silently watching and listening to the forest and its story. I saw birth. And I saw life. And I walked the summer trail with the wolf. My days at Source Lake---were like pages from a book. Over the months---the and complete. It was a story of the wildlife Icmme to know---in a land still untouched by man. story became clear It had a beginning---and it had an end. But the end was a new beginning. A new story. This film is but one of the stories from Source Lake. It's a story of the land as told from the past. ‘ I felt---that if the forest could tell its own story---if it could speak---it would do so through the wisdom of the Native American. He used to dwell there. He was close to the land. But he's now gone from the forest---and with his passing---we have lost a source of wisdom. But---if his voice could be heard. If he could once again speak for the land. If he could speak to the children. If he could speak to you. This, then---is what he would say. I am Anishnabe. I am the forest. Native/grass Hawk/sunburst Turtle Island Huntsville Swamp Caribou herd Bull caribou Carribou herd Wolves Wolf pack Native 192 My talk is the voice of the wolf---bear--- procupine---and deer. I am the mountain meadow---I am the ground hog. Marshbird and sweet grass---I am these things too---and more. The spirit who you call the Creator or Nature---made all that is. All went well on the land. The circle was complete. The sun gave light and warmth and the plant beings responded with growth. Some grew to meet the light---others held fast to the rocks.’ Slowly the soil was made. The rain that fell---was gentled by the plant beings---and the new soil held. The lakes filled with rain and beings began to live there, too. Seeds blew on the winds---and traveled on the backs of beings. The forest moved to new places---but it stayed here, too. Every being, then---was a part of the circle. It was perfect. Creatures grazed on the plant beings---and their numbers grew. Yet the plant beings were not wiped out--- for hunters stalked the land---predators to keep the others alert and sharp. These mighty hunter beings---exists as a sweet song to the circle. For they bind it together. They and the forest as one. Together it is complete. But---take them apart and it's over. Listen, children---so I can tell you of a great predator. Wolf being---our brother. I tell you this before the shadow of trouble that walks the 1and---wipes him out for good. Wolf walk Beaver Wolf Native Forest Moose Lone wolf Native Wolf Forest Caribou bull Duck flight 193 Already---he has suffered much. His trail is now in sunset. Before your people came to this land--- wolf being was. He and the others were working in all the places. You see---the circle was perfect. It could be left alone---and it kept on working---all by itself. As you knew right from wrong---I speak of the forest as right. There all was right. Wrong and bad---they are your words. They are not found in the forest. All beings in the circle served a purpose. Each was a part---together they were the whole thing. Wolf being never was a bad guy. But many of you do not believe this. Amongst your people---he is known by your word---because he kills to be a part of the circle. But he is not angry or crazy. Like all the others---he, too, must eat to live. To continue. He does his part in the circle---and he does well. You see, children---before the circle could be as it is---wolf being had to be. Predator---keeper of the forest. He keeps the circle strong. Wolf takes the beings from the forest--- but this must be done. The forest is rich---plant beings provide for all. There---creature beings can live real good. But When their groups grow too large---and they get fat---or sick---or don't watch out---wolf being takes them. By this you see---wolf keeps all groups alert. MOOSE Wolf Forest Pack wolf Buck Wolf Bull moose Wolf Running moose Wolf pack Female/pups Marka Marka/Woody Pack Marka/stare Deer herd Wolf Wolf Bear Wolf Bear 194 You have to be sharp to get by in the forest. Wolf makes sure there are just the right number in each group. And because of this---the forest can carry all groups of beings. To do his work in the forest---wolf had to be special. He had to be smart---to use different ways to hunt. For the deer was aware of his presence and would not wait for an attack. He needed great courage---for a big bull moose would be glad to kick him to death. He had to be swift---and be able to take a being from the circle quickly.- But wolf is also a social being. He lives with others of his kind. Together they are the pack. Although their work is to take beings from the forest---in their group there is peace. They laugh and play. They give gifts to each other and they play tricks, too. They provide for each other. And they talk---not as you know talk--- but talk just the same. They communicate and work together. They are the hunting pack. Together---they are strong---the circle is complete. So great is the wolf---that he even allows herds of deer to go unhunted for years so that their numbers may increase. You call this farming and are surprised. For the forest and the wolf---this is the natural way. It is no big thing. No hunter being preys upon wolf. He has no enemy there in the forest. It is true that the bear and he do not get along---but there is no hate. When they meet---each tries to chase the other away. Wolf Forest Scenting wolf Forest Marka howl Forest Marka/Woody PupS' Pup run Female/pups White wolf Woody Pack wolf 195 It is equal. But no being in the forest is without control. For the wolf this is also the way. It seems hard---but it must be. In the circle---it is the way. But wolf is a great being. Constantly sensing the forest---he seems to know. He knows some of tomorrow in the forest. Very smart fellow that wolf. Remember---I have spoken that the forest can carry only so many in each group. And that this, too, is for the wolf. But he knows---his body responds to all the forest sensings. Wolf gets ready for things to come. Whether new pups will be born to the pack depends on how well the forest is doing. Most times only the top two in the pack have pups. They are the leaders---the strongest. They are made of the very best stuff. So, too, their pups will be strong---and the pack will remain sharp. You have to be sharp to get on in the forest. But more than one-half of the new wolf pups don't make it. Things can change fast in the forest. Like the spring flowers---wolf pups are delicate. Spring flowers and wolf pups often die early. Those who make it---those who join the circle as big wolves---have troubles to face, too. Like their human brothers---they get many of the same sicknesses. And they die. Wolf himself is a forest for small creature beings who live on and in him. Moose Lone wolf Cliffs Swamp Woody scenting Marka Owl Marka Pack wolf Woody Pack wolf Swamp Deer Native Cliff 196 Here wolf is helpless. It is not in his control. And during the hunt---the moose can kick him crazy. Being a wolf is not an easy life. But enough get by. In the forest it is the way. The circle continues. Because the forest is so great---it takes many wolf packs to keep the circle. So the packs would not all gather up in the same spot---they marked off territories. Parts of the forest where only one pack would hunt. Wolf sets up a barrier all along the edge of his territory. A trail of smells---special scent messages---telling things like---where you are---who's around---and what's going on. The scent marks are from what you call body waste. But in the forest there is no waste. Scent marks are strong messages. Other wolf packs show respect---and they will not cross over them. If they do and they are caught---they will be killed most times. Scent marks and territories are big stuff. They help keep peace in the forest. They keep neighbor packs from fighting each other. Each pack stays within its own territory ---and there is no trouble. For a thousand years and more---that is the way the forest was. For all that time the circle was strong. And while it was strong--- I, Anishnabe was living there. I understood that I was part of the circle ---that I was bound to it. And I did not struggle to free myself. All went well on the land for a long time. Cliff Pack wolf Nursing wolf Pup run Native FOY‘ESt City Forest Native Forest 197 Then the white man's tribe came to the forest and things began to change. The new beings looked at the mother forest and saw it as their enemy. But it never really was. They struggled with it and tried to defeat it. They slaughtered the wolf---and drove him from the forest. Where there once were one hundred wolves in number---there now remains only one. All the rest are wiped out. It is sad. The white man's tribe never knew the wolf Instead, they let their minds spin with wild and crazy thoughts of some wicked beast. They made the wolf into something it was not. Then believing their own imaginings---they killed the wolf. Children---I worry about the circle. I am afraid that like many who came before you---and are here now---that you, too, might turn your back upon the mother forest. I fear---that you may occupy your minds with thoughts of breaking with the land. Now---today---many among you go for days without placing their feet upon the land. Many stay up on the concrete platforms and blacktop trails---within your villages. Many think away from the land. Yet---they can never go away from it. The land is our life---without it---we are nothing. Children---before my voice returns to the forest---and I speak with you no more forever--- I ask that you never turn your back on the an . Always remain close to it. And remember---each life has a reason. Female/pups Marka/howl Credits 198 All beings are part of the circle. Together---we are the whole thing. Grow with a respect for all this---and you will have done well. The circle can be strong---once again. APPENDIX I APPENDIX I MODEL: PERSUASIVE MESSAGE DESIGN 199 200 s_ u :8 gig—mg 2‘95- 822 8 5300.5 3:: g 35:5 995—: C 2.3.- 0953 . a: :u- ou~§ ¥_‘l ’- in: o C 5.... = 3.6 7 Juan“. 2....» mm mm min mm 5 2.82 .I .285 201 advocated (“IUSIM Acceptance of Directed I! VIII“ Olmr crélpfl: Cy Illa I- ' MCI 1" m ., I l I I 1 I 1 I I 1 I 1 1 I I I l 1 I I 1 I 1 202 . . . . . 58:96 . L 85. 3 8 2.8 2.8 8... o _ air ... A a: a. ism-er. 3F. 88 I . . . _ _ . _ . as! .8 5589... 8. .51.! u a. 3:3 8: FEW.” " - - \l . . .m _ 1—2- 83 i IIIIIIIIIIIIII new... w .84.... age». are... . l a u L . _ .,. 85...... 25... . F 11111111111111 I. u .m o 8 2 . . ... . 8.88 o ... _ .... an: . $3.. _ _u a _ _ _ _ . _ _ m I 838:. u . u. . . . m an a: 5... I. it . ... _ . A _ u . . 88.0 8 :8 . ..8 I _ I: . :- . an... awe. . insist .....a... ......E... u _ . r IIIIIIIIII L n . rillllllllIIIIL alas. Ab 203 38.8.. c z- ! - 5:83.... 8.”...8 fl . 8%....- fiu): . . - O ... g. 38. I i .8 g o» 8 85. J ”38.8 = .8 888- a... 5:8. .858 8.2.38. 83:8 2'» oz... duh... 8o ... sac! ...—Junk“... figs—9.. u .88 9%.! e...» I ' > is if». . :- 2 3 3.28.5 25!...- E . :3: .858. 8% . ......— nno... u .88 > 1 F8...- uoll... 3.8. 8- . 8.. 32:8. .98 993.6 :- .. ...-...: ....3 gig. 53:5 8289 In...» fig II :4- 8“... ...—.630“; f 88.... 8.58. . .3388... . :88 88.90%. .3828. _Il 24%-h"? 329:8 «88.8. ...-3.88... 3:3: a 58...: o 888. :8 888- a! 1 8. 2.838.. 8.... .... 2:83.... .2 aux? I.“ 204 . . .8 8 8. . llllllllllllllllllllllllll c — Luuéhflcw .* H “CH-:— . _ .2 38 858 ...-.8. I. . .8 .. o 3.... .3. . 83 ... . . «3 .m .8. 8... u . v.93»... .... . ) 38- 88. ow... .m .n I - ' 8.3 88 E 8.838 .m _m z . ... I ...-.8... a ...... ......1 ”when ll 73. ... ...-SI . .m 85 8.08.“ .2 «“98: .m .. _l\ .l. . — . 1 S38 . . .838 9.8 ha #8 8.3... .... 3... _ ...-:3 o .8 889.8 . .838 .. .38 a 3 ”“8 .... 5”“..- .8:8"8_8 . l 4 n . .838... ..r . _ . _ . . . . 2....- .. .83.... . — u 50 ,. § . ....E... H .2... .M .... 85.838 0.....- a... . . . ''''''''''''''''''''''''' - pi-"InI'-IIIIII-'|"'-|'I"-'I|'II- 53! £— 205 9.38.8. 093.2. *0 5.2.8... 8. “'50:: 28388.53... Co..uml..8.u>8 3 “is a; on: . ...».n.......%..u 933.3. 8.... ...8 8861 885 3 >28 2.8 838 . .823 m .3 8.3.8.. Goya; “a“: _.ugb—>s .Bo . Co. 8 on 8.3.. .... 32.8 5... 855 .2 ”88.38 ... .8 3.5. :8 .8 3. 8.82%- .--- — _ _ --(-SJ(;33;J-=ImTal-D'NT_—-- «ll-III- '.l'||l|||l. can! 0:350; pug .> u .