Influence of Motion Pictures on Children's Attitudes
Louis L. Thurstone
Psychological Laboratories of the
University of Chicago
This is one of a series of studies to ascertain whether the effect of motion pictures on school children can be measured and whether the effects of different kinds of pictures can be predicted. The present report concerns two experiments with the films "Street of Chance" and "Hide Out." The plan of each experiment was to ask the children to fill in certain schedules intended to reveal their attitudes on the issue which seemed to be involved in the film. These schedules were filled in by the children before and after seeing the film. If the film had any effect on social attitudes it might be revealed by changes in the scores on the schedules filled in before and after seeing the film. Such effects have been found to be measurable for several films.
"The film "Street of Chance" describes the life of a gambler in such a way that the children might conceivably be affected in their attitudes toward gambling. By special arrangement this film was shown in the Strand Theater in Mendota, Illinois, on May 22, 1929. Mendota is a town of about 4000 population. There were 240 children in the experimental group in Grades 9 to 12 inclusive. A paired comparison schedule of crimes was presented to the children on May 15, 1929, before seeing the film, and also on May 23, 1929, after seeing the film. The second filling in of the schedule was done on
( 292) the morning after seeing the film. The children were given free tickets to the local theatre where the film was shown in an afternoon performance. This was done in order to give as much as possible of the natural setting for the effectiveness of the picture as they are ordinarily seen by the children.
The paired comparison schedule had the following instructions :
A STUDY OF ATTITUDE TOWARD CRIME
Write your name here
Boy or girl Age Grade
This is a study of attitudes toward crime. You are asked to underline the one crime of each pair that you think should be punished most severely. For example, the first pair is:
speeder — pickpocket
If, in general, you think a speeder should be punished more severely than a pickpocket, underline speeder. If you think a pickpocket should be punished more severely than a speeder underline pickpocket. If you find it difficult to decide for any pair be sure to underline one of them, even if you have to guess.
speeder—pickpocket bank robber-gambler
drunkard—beggar quack doctor—bootlegger
Then followed 78 comparisons of the type indicated above. There were 13 crimes in the list and every crime was paired with every other. The children were told that-they would be asked to fill in these schedules twice so that they would not be surprised when asked to do it a second time. The schedules were filled in during school hours and the picture was shown in the local theatre. Nothing was explicitly said about any connection between these two events, but there is no guarantee that. some of the children may not have suspected that there was an association between them. The ideal procedure is to separate these events so that the subjects do not think of any association between the picture to which they are given free admission and the attitude schedules which they fill in during school hours.
In Table 1 we have a summary of the raw data for the schedules that were filled in before seeing the picture, and in Table 2 we have a similar table for the schedules that were filled in after seeing the picture. Table 1 shows, for every pair of offenses, the proportion of
(295) the children who thought that the offense listed at the top of the table more serious than the offense listed at the side of the table. If the proportion Pa>b is very high, say .90, the interpretation is, of course, that the children thought generally that offense a is much more serious than b and that a should be the more heavily punished.
With thθse data it was possible to calculate the scale value of each o the thirteen offenses by the law of comparative judgment (2, 3, 4, 5 ,6, 7). The simplest form of this law was used for these calculations, namely, Case V, so that
Sa Sk = xak √2 (1)
in which Sa and Sk are scale values, and xak is the deviation from the mean of the probability surface in terms of its standard deviation which corresponds to the observed proportion of children who said that a was more serious than k. In the same manner we may write the equation for the two stimuli b and k in the form
Sb Sk = xbk √2 (2)
Subtracting from , we have
Sa Sb = √2[xak xbk] (3)
Writing this equation in summation form so as to involve all comparisons with the stimuli a and b, we have
n(Sa Sb) = √2 [xak xbk] (4)
(296) from which it follows that
Sa Sb= (√2/n)[xak xbk] (5)
This is the equation used for calculating the scale separation between the two stimuli a and b, and similar forms were used for the calculation of all other scale separations. The numerical values of XX and of Xbk were obtained from the tabulated proportions in Tables 1 and 2. Since the details of these calculations have been previously described, they will not be repeated here (4, 7).
The first column of Table 3 gives the' names of the offenses in the list. The second and third columns give the scale values before and after seeing the film "Street of Chance." In Figure 1 we have plotted the scale values "after" against the scale values "before." A linear plot is immediately apparent with a conspicuous exception for the scale value of gambler which was evidently rated as much more serious, relative) to the other offenses, after seeing the film. Low scale values represent the more serious offenses. The variations in the "before" and "after" scale values for each of the other offenses show very slight changes which may be interpreted as due to slight chance errors in the experimental proportions.
The shift in the scale value for "gambler" cannot be so interpreted. The film quite evidently had the effect of making the children regard gambling as a much more serious offense than they did before seeing the film. The same effect is shown graphically in another manner in Figure 2. Before this figure could be drawn it was necessary to adjust the scale values to a common unit. The adjustment was made on the "after" scale values. The nature of the adjustment can be explained as follows.
Imagine that the subjects were asked to fill in these paired comparison schedules one hundred times. No matter how much interest they have in the task or in the issue involved, they would get bored with the performance until their underlinings would finally become so indifferent as to be almost a chance matter. They would, nevertheless, still regard certain offenses as more serious than certain other offenses, but their indifference to filling in the schedules would obscure the affective values of the stimuli. It will be noticed that the slope of the linear plot in Figure 1 is not unity. The spread of the scale values on the second occasion is slightly smaller than the spread of scale values on the first filling in of the schedules. In fact, the slope of the line in Figure 1 is .95, as determined by the method of averages. This deviation of the slope below unity is a measure of a slight degree of indifference to the task on the second occasion as compared with the first. The discriminal error was slightly larger and the scale separations in terms of the discriminal error, therefore, seem to be slightly smaller on the second occasion. The adjustment of the second set of scale values is made by assuming that the average true scale value remained unaltered by the film and by the filling in of the schedules. A stretching factor of .95 was introduced into the second set of scale values so that the two sets would
( 299) be directly comparable and these two sets of scale values are shown graphically: in Figure 2.
If Figure 2 were drawn without this adjustment for the slight enlargement of the discriminal error the second set of scale values would have a spread slightly smaller than the first. In the present instance, the. scale value of "gambler" made such a large jump that the comparison with special regard to this one offense would be practically unaffected by the adjustment. However, theoretically, the adjustment should be made in order to take account of the fact that the discriminal error is increased by a slight amount of indifference 6r boredom with the repetition of the task of filling in the schedule.
In Figure 2 we have the two sets of scale values placed in proximity for direct comparison. It can readily be seen that the scale values do not change markedly except for "gambler" which was regarded to be a much more serious offense after seeing the film than it was before the performance. The film "Street of Chance" was selected for this experiment because it was thought that it might even make the children more lenient toward gambling, owing to interest in the gambler who was the principal figure in the film. The results of the experiment show dearly that such was not the case. The film had the opposite effect, namely, to make the children regard gambling as more serious than they did before. It is also quite likely that the concept of gambling was altered in the minds of these children. It is not unlikely that many of these children were unfamiliar with high-powered gambling and that the film caused a shift in the cognitive aspects of the concept of gambling as well as in the affective aspects. On the whole; the film may be said to have a socially approved effect on the children because they became more severe in their judgment of gambling even-though a gambler was an interesting principal figure in the film.
Another experiment was carried out at Princeton, Illinois, with the film "Hide Out." This experiment was planned in the same manner as the previous one. The film was shown in the Apollo Theater in Princeton, Illinois, on May 26, 1929. This town also has a population of about 4000. The film was seen by 254 school children in Grade 9 to 12 inclusive. The schedules were filled in during school' hours on May 19, 1929, before seeing the film and again on May 27, 1929, the morning after seeing the film. This film was a
( 300) second choice for the experiment since it was found that our first-choice film for this experiment had already been shown in this town-The experiment was completed although we were. not satisfied with the film for our present. experimental purposes. It involved boot-legging, and we decided to measure attitude toward bootlegging in two ways. The paired comparison schedule of offenses -was givers to the children before and after seeing the film. ' Bootlegger is included in the list of 13 offenses, and it was thought that attitudes of the children toward bootlegging might be affected by the film. We also gave this group of children a statement scale about prohibition. The instructions for this statement scale about prohibition were. as follows :
A STUDY OF ATTITUDE TOWARD PROHIBITION
Write your name here
Boy or girl` :.: Age Grade
This is a study of attitude toward Prohibition. On the other side of this page you will find a number of statements expressing different attitudes toward Prohibition.
√ Put a check mark if you agree with the statement.
X Put a cross if you disagree with the statement.
Try to indicate either agreement or disagreement for each statement. If you simply cannot decide about a statement you may mark it with a question mark.
This is not an examination. There are no right or wrong answers to these statements. This is simply a study of people's attitudes toward Prohibition. Please indicate your own convictions by a check mark when you agree and by across when you disagree.
Before each statement in the following list we have recorded its scale value.. These scale values were, of course, omitted from the printed schedules that the children filled-in. The scale values were used in determining the attitude score of each child toward prohibition. The detailed methods of constructing these scales have also been previously described so that they need not be repeated here (4).
√ Put a check mark if you agree with the statement.
X Put a cross if you disagree with the statement.
5.5 1. It is absolutely immaterial whether we have prohibition or not.
8.6 2. The Eighteenth Amendment should be repealed and local option adopted.
6.4 3. Prohibition should come as the result of education, not legislation.
.8 4.The entire state and national resources should be mobilized for prohibition enforcement.
8.2 5. Liquor `should be sold by licensed liquor dealers in restricted amounts.
1.4 6. Prohibition should be retained at all -costs.
9.3 7. Prohibition is undesirable because it drives the liquor traffic underground rather than eliminates it.
11.4 8. Possession of intoxicating liquor in any form should subject individuals to punishment.
9.2 9. Prohibition should be a matter to be decided by the individual, and not by the government..
3.5 10. The present prohibition laws are necessary for the good of the United States.
7.5 11. Manufacture of wines and beer in the home should be permitted.
10.4 12. The open saloon system should be universally permitted.
6.9 13. Prohibition is not desirable now because there is not a sufficiently large-majority in favor of it to make enforcement effective.
5.6 14. Both good and bad results have come from the Eighteenth Amendment
10.2 15. Prohibition has been tried and has proved a miserable failure.
3.7 16. While the Eighteenth Amendment is a part of the constitution it should be observed.
7.0 17. Prohibition is good in principle but it is doing more harm than good because it cannot be enforced.
10.2 18. The Eighteenth Amendment should be repealed.
2.5 19. Prohibition prevents many accidents and should, therefore, be enforced.
4.6 20. It must be admitted that the Eighteenth Amendment is a restriction of personal liberty, but it has benefitted many people.
3.2 21. The national government should increase its appropriation for prohibition enforcement.
4.4 22. Although not completely satisfactory, the present prohibition is preferable to no prohibition.
3.3 23. The restriction of personal liberty under prohibition is entirely justified by the benefits.
4.5 24. The experiment of prohibition may prove to have some value and may, therefore, be worth trying.
7.8 25. Prohibition is an infringement upon personal liberty.
2.4 26. The effect of prohibition on the national life of America is more than constructive.
2.3 27. The present prohibition laws are satisfactory and their enforcement should be more severe.
1.8 28. Since the liquor traffic is a curse to the human family it must be dealt with bylaw.
These statements were adapted from a doctor's thesis by Mrs. Hattie Smith (7).
In Figure 3, we have plotted the "after" scale values of the 13 offenses against the "before" values. The diagram shows a linear plot with no conspicuous deviations, and hence we conclude that the 'affective judgments about these offenses were not affected by this .film. The slight variations are due merely to chance fluctuations in
( 302) the experimental proportions. In Figure 4 we have plotted the frequency distributions of scores on the statement scale of attitude toward prohibition before and after seeing the film. The two means are indicated on the base lines of the distributions. This diagram reveals also no significant change in the attitudes of these children toward prohibition as a result of seeing the film "Hide Out." There was a slight change in the mean scores slightly more favorable toward prohibition after seeing the film than before seeing it, but we do not regard this change to be large enough to be attributed to the
( 304) film with any degree of certainty Our conclusion is, therefore,' that the film "aide Out:'did not have any measurable effect on the attitudes of the children toward bootlegging or toward prohibition.'
These two experiments illustrate the application of two methods of measuring
affect, namely, the paired comparison procedure and the statement-scale
procedure. In other experiments both of these procedures have demonstrated
measurable effects of motion-picture films on the social attitudes of school
children. In one of the films here. discussed, namely, "Street of Chance," a
conspicuous effect of the film was demonstrated. The film made the children more
severe in their j figment of gambling than they were before seeing the film. It
seems to be evident from these experiments and from others of a similar type
that motion pictures can be used to affect the social attitudes of school
children and that these effects' can be objectively measured.
1. Smith, H. N. The construction and application of a scale for measuring attitudes about prohibition. Unpublished thesis, Univ. Chicago.
2. THURSTONE, L. L. The method of paired comparison for social values. J. Abn. & Soc. Psychol., 1927, 21, 384-400.
————. Psychophysical analysis. Amer. J. Psychol., 1927, 38, 368-389.
————. A law of comparative judgment. Psychol. Rev., 1927, 34, 273-286.
————. A mental unit of measurement. Psychol. Rev., 1927, 34, 415-423.
————. Attitudes can be measured. Amer. J. Sociol., 1928, 33, 529-554.
————. An experimental study of nationality preferences. J. Gen. Psychol., 1928, 1, 405-425.
University of Chicago