Measuring the Influence of Motion-Picture Attendance on Conduct and Attitudes
Frank K. Shuttleworth
Out of one hundred children in the junior high schools of large urban centers approximately twenty-seven attend the movies two or more times a week. Seven go three or more times and two go four or more times a week. What are the movies doing to the conduct and attitudes of these children ?
When this question was originally raised by the Motion Picture Research Council, the experimentalists at once proposed the following procedure : First, select two large groups of children alike in as many respects as possible, one to act as a control and the other as an experimental group. Second, measure both groups by some objective test of conduct or attitude. Third, subject the experimental group to a motion picture which contains promise of influencing the measured conduct or attitude. Fourth, re measure both groups and see if the scores of the experimental group have changed more than could be accounted for by chance. The studies by Thurstone of the influence
( 217) of specific movies on specific attitudes constitute an excel-lent example of the precision of this approach. Given adequate tests and care in handling the actual execution of the experiments, it is obvious that the results are clear and unambiguous : exposure to specific movies either does or does not change specific attitudes.
It was equally obvious, however, that such an approach would fall short of meeting the real issue. The complaint against the movies is not that specific films influence specific conducts and attitudes, but rather that the general run of movies has a generally unfavorable influence. To test the influence of the general run or of a random sample of movies is something very different from testing the influence of a specific movie which has been selected primarily because it promised to exert a certain influence. Further, to measure generally unfavorable or favorable influences would require an enormous range of tests in a field where adequate tests are few and far between. The study conducted by Professor Mark May and the writer was an attempt, in part, to solve these difficulties.
Our procedure involved three steps. First, the selection of groups of children who go to movies frequently and groups who go infrequently. Second, the equating of these selected groups for as many other factors as possible. Third, the comparison of the selected frequent and infrequent movie attenders on a wide variety of tests of conduct and attitude. All told 516 frequent and 543 infrequent movie-goers were selected for study from among nearly 6,000 children in grades five to nine. These selections were based on the children's own report of their movie habits. The reliability of these reports is at least .60 and possibly .70. Throughout, the two groups were equated for sex, age, school grade, intelligence, and socio-economic educational home backgrounds. The first comparisons between movie- and nonmovie-goers employed the conduct, reputation, moral knowledge, and attitude test-data collected by the Character Education In-
( 218) -quiry.[1] Here 102 frequent and 101 infrequent movie at-tenders selected from among nearly one thousand children were studied intensively. Specially constructed attitude tests were given to 106 movie- and 102 nonmovie-goers and a revision of these tests was given to 308 movie- and 340 nonmovie-goers under conditions which led the children to believe that their responses were anonymous. The revised attitude tests contained 343 test elements which were designed to measure the influence of seventy-one carefully defined attitudes. The test elements consisted of true-false statements, multiple-choice questions, and a wide variety of other devices for eliciting attitudes. The evidence is that children's responses to such questions are to a substantial degree their own independent answers. The analysis of the attitude tests was in terms of the individual test elements. While the reliability of a single test question is not high, averaging only .34, several questions were directed towards each attitude, large numbers of children were involved, and the contrasts between the movie and nonmovie children are extreme.
These procedures yield about a hundred reliable or nearly reliable differences between frequent and infrequent movie attenders which may be grouped into thirty-seven tendencies. With few exceptions the frequent movie-goers make a poorer showing on the conduct tests and display less desirable attitudes than do the nonmovie-goers. The nature of the differences, however, makes it very doubtful whether they can be attributed with any assurance to the influence of the movies. Only four of the thirty-seven tendencies can be traced directly to the movies, while twenty-four may be attributed in part to selective factors. For example, the movie children tend to affirm while the non-movie children tend to deny the following statements: Most policemen torture and mistreat those suspected of crime; few criminals escape their just punishment; most
( 219) Spaniards are impractical, romantic, and love makers; few Russians are kind and generous. Examples of differences which are probably due in part to selective factors are the following: Movie children receive poorer deportment and scholastic marks and are less interested in school; they are less coöperative, less emotionally stable, less honest in school situations while equally honest out of school; they are more interested in cheap reading, in dances, in a thrill, and in fine clothes; they appear to lack inner resources for keeping themselves busy and entertained. Such children would naturally gravitate to the movies. On the other hand, children who are interested in their school work, who are practical and serious minded, and who are busy with other activities simply do not care about the movies. We fully anticipated that such selective factors would be involved. The point of these examples, however, is that diligent search for differences which could be attributed to the movies was meagerly rewarded, while a systematic study of the data of the Character Education Inquiry in which we hardly expected to find differences has revealed many which appear to be due to selection. Instead of measuring the influence of the movies, our results serve almost as well to define the characteristics of children who are attracted by the movies. Probably excessive movie attendance serves to stimulate and aggravate these characteristics, but whether this factor or the factor of selection is more important cannot be determined.
