commentary, scheduled programs such as "Science and Engineering," "Viet Nam Fights Back," "What is Communism," "Moscow Mailbag" (with its inimitable slogan, "You couldn't do better than write us that letter"), and short serials such as "The Pechala Family."

Several characteristics of the programing are particularly important. First, content analysis indicated that most of the programing did not deal with international political affairs per se. 17 Aside from short news items, some political commentaries, and a few regularly scheduled programs on political issues, the broadcasts dealt mainly with such topics as urban planning in Moscow, descriptions of the educational system in the USSR, grain production, the rebuilding of the city of Frunze, the booming concrete and cement industry, and the excelling of work norms in tractor production-all quite descriptive and factual in character. Second, by both content analysis and a panel of 15 judges, the programs were evaluated as "moderately biased."18 These judges also rated the manner of presentation as "dull, boring, and unimaginative."

One striking exception to both of these observations, and one of particular significance in the analysis of results, was the issue of Viet Nam. The subject of news shorts, political commentaries, and a weekly feature of its own entitled, "Viet Nam Fights Back," this issue was frequently presented in a blatantly one-sided manner, characterized by highly charged negative statements likely to be viewed as incredible by an American listening audience, such as reports of U.S. troops ramming flaming bamboo splinters underneath the fingernails of captured prisoners, or U.S. troops pushing captured Viet Cong out of high-flying helicopters without parachutes. 19

Several separate indices were used as indicators of the subjects' perception of the broadcasts in comparison with their expectations. The results reported here are based on the subjects' replies to questions in the after questionnaire which asked whether or not the content of the broadcasts, and the manner of their presentation, were what they thought they were going to be before they listened and, if different, how and in what way they were different. These items provide the basis for categorization of the subjects into those individuals who perceived what they expected (referred to hereafter as the expected group), those individuals who perceived the broadcasts to be worse than they expected (the unexpectedly worse group), and those individuals who perceived the broadcasts to be better than they expected (the unexpectedly better group).

RESULTS

Data are given here on the following items: the Soviet government (evaluated by semantic differential), separate measures of the subjects' views of the Soviet government's sincerity, the legitimacy of that government's actions in world affairs, and that government's actual desire for world peace (each item evaluated on a nine-point opinion scale), the Russian people, and American military participation in Vietnam (the

latter two each measured by semantic differential). 20 Since the subjects were continually exposed to the many reinforcement patterns of their own society while being exposed to Radio Moscow's broadcasts, we could hardly expect actual conversion of their opinions; consequently, the measurement here is of the direction of opinion change rather than conversion of opinion.

The initial results are given in Table 1. Even without any breakdown within the experimental group, it is clear that the broadcasts did have some impact.21 In Table 1 there is no significant difference between the opinion change occurring in CI and that change occurring in CII, an indication that participation in the experiment was not a significant factor influencing the opinions of the participants. On the other hand, there is a significant difference between the experimental group and CI (in the case of the desire item, CII) on all issues but Vietnam, an indication that the broadcasts did indeed result in considerable opinion change.

Table 2 shows the data bearing on the specific hypotheses of this report. Hypothesis 1 stated that those listeners who perceived the broadcasts to

For purposes of this table it is assumed that opposition to the U.S. military participation in Vietnam would be favorable to the USSR.

be relatively better than their negative expectations would respond positively to their exposure. In Table 2 it may be seen that this unexpectedly better group consistently changed in a direction more favorable to the Soviet Union on all issues except Vietnam, an issue which, as we have noted, possesses considerably different characteristics. The change is particularly striking on such issues as sincerity (70 percent of the subjects became more favorable, while only 8 percent became more unfavorable), and on the semantic differential of the Soviet government, on which 63 percent of the subjects moved in a favorable direction and 7 percent became more unfavorable.

In contrast, on Vietnam, the one issue which was clearly dealt with in the "hard line" fashion the groups had expected, the unexpectedly better group did not move particularly toward the Soviet Union. Although we have no separate measure of their perception of this specific issue, subjects in the unexpectedly better group frequently indicated that on this one issue, they perceived the broadcasts to be as negative as they had expected.

Hypothesis 2 stated that those listeners who perceived their negative expectations to be relatively confirmed would be negatively affected by their exposure. In Table 2 the data are mixed for this hypothesis. On the issues of desire and legitimacy, individuals in this expected group did move in a direction more unfavorable to the Soviet government. 22 However, they split evenly on the issue of the government's sincerity, and, on the measures of the Soviet government and the Russian people (both obtained by semantic differentials), they actually became more favorable (although certainly not as much so as the unexpectedly better group). These results suggest that under some conditions subjects who generally perceived the broadcasts negatively can be positively affected. It may well be that, just as the unexpectedly better group perceived the Vietnam issue differently from other issues, the negative expectations of the expected group were not confirmed on these specific issues. In future research separate measures of expectation should be obtained for each specific issue.

Hypothesis 3 stated that those individuals who perceived the broadcasts to be relatively worse than their negative expectations would respond most negatively. Although the N in this group is so small as almost to preclude the results from reaching a .05 level of statistical significance, opinion change in a direction more unfavorable to the Soviet Union is, as hypothesized, clearly the most pronounced among subjects in this group (see Table 2). Individuals in this unexpectedly worse group consistently moved in a direction negative to the Soviet Union on all issues except sincerity (on which they split in the direction of their opinion shift). 23 On the issue of desire, for example, 46 percent of the group moved in a direction unfavorable to the Soviet Union while 15 percent became more favorable. Even on the item of the Russian people, about whom all categories in the experimental group originally had rather favorable

Table 2

Percentage of Subjects Modifying Opinions Toward the USSR, by Prior Expectations

Due to the limited number of cases, the statistical test used for this group is the binomial test, one-tailed.

opinions, the unexpectedly worse group responded quite negatively, with 46 percent becoming more negative, and 23 percent more favorable.

The subjects' open-ended replies on the after questionnaire flesh out the results of these before-after measures. Some excerpts from the unexpectedly better group illustrate the subjects' reactions:

[1] A very silly effect, perhaps, but all of a sudden I realized that Russia's citizens are ordinary people.

[2] You know, here we are always cast in the good guy role. It (the broadcasts) helps you to step back and get a possible view others might take of us and their reasons.

[3] You begin to see that the Soviet Union is not a monster with atomic bombs in each hand; instead they're human, as concerned with human affairs as we are. [4] Through the dark and heavy mist, every once in a while I felt they really did seem to want world peace, at least those in control of the broadcasts, and as I look over the world situation as a whole, I can begin to see where this could be

true.

Such replies are very similar to those we consistently get in interviews with actual American listeners to Radio Moscow's North American broadcasts. Frequent comments from the "real-life" audience are: "It makes you realize they are people just like us"; "You see that there are two sides to most of these international problems"; "You begin to see their side of the story too." 24

The open-ended responses also provide additional support for the theory entertained in this research-that the process involves a comparison of what the respondent heard with what he expected to hear: [1] I expected it to be one mass of making the U.S. look black and completely positive for the Russians. I thought I would hear more bad points than I did. I was surprised, they seemed so much like us. From the general idea we get in America we are told the Soviet Union is bad, so I assumed they would try to make us look bad too. They were fairly reasonable about it and it made me think maybe they really were trying to help us understand each other better. [2] You can't believe all they say, but they didn't hit us hard on the things I thought they would, like race riots, as if they really were trying not to make trouble.

[3] I thought they would try to persuade me. Instead they played music and frankly answered those questions from Americans. They seemed so sensible about it all-makes you realize they're honest people trying to help international affairs.

There are undoubtedly many sociological and psychological factors at work to account for why some subjects perceived what they had expected in the broadcasts and some did not.25 In our research so far, we have examined only one of these factors-dogmatism, with the more dogmatic individuals purportedly more resistant to opinion change and more prone to distort information to fit pre-existing opinions and beliefs. 26 There is some indication in these data that those individuals scoring high on the dogmatism scale perceived the broadcasts as they had expected them to be, while the low dogmatics found the broadcasts to be something different from what they had expected (t= 1.37, p <.15).27

DISCUSSION

This study obtained a measure not only of the specific point on an opinion measure which a subject indicated most exactly characterized his