페이지 이미지
PDF
ePub
[ocr errors]

current knowledge about the legislature, measured at T2 only. The maximum score was 16 and the average score for all respondents was 8.7, or an average score of 55% correct in terms of current legislative activities.

Watching the television series led to average knowledge scores of 9.84;

the non-viewers in the experiment scored 7.64 (Table 4).

In addition to the

statistical significance of this difference, it is important to note that

Again, to look at the possibility that exposure improved moreso the

know ledge levels of those already more knowledgeable, we compared the results

on this index between black and white youngsters. Within the control group,

there was no particular difference in level of knowledge between blacks and

whites. However, the white youngsters in the experimental group correctly

answered two more questions on the average than the black youngsters who had seen the series. In terms of the kind of differential learning rate being discussed, the white viewers answered correctly 2 1/2 more questions per person than the white non-viewers, whereas the black viewers answered correctly one more question than the black non-viewers.

The final measure of knowledge dealt with the individual's understanding of how the legislature tends to work. For example, what do amendments do, the importance of committee hearings, the attention given by legislators to floor activity. The maximum score on this index could be 23. The average score for the viewers was 16.36. This compared to an average score of 12.31 for the non-viewers or control groups. Obviously, watching the series increased information at put the nature of the legislative process. Again, the overall level of effect was to increase by 1/3 the available knowledge

among all the viewers.

Table 4

Effect of Program Viewing on Current Activities and Structural Knowledge

[merged small][merged small][merged small][merged small][merged small][ocr errors][ocr errors][ocr errors][merged small][merged small][ocr errors][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]
[ocr errors]

We looked again for any evidence that such knowledge increments might

be a differential function of the original knowledge level of the viewer,

and that this in turn could be related to one's social background.

The white

respondents in both the experimental and control groups scored significantly higher than their black peers. The difference in knowledge between the white viewers and non-viewers was four units; the difference between black viewers and non-viewers was three units. Although this is a smaller gap, it repre

sents fully 25% more learning among the presumably more advantaged learning

group.

For all these knowledge gap comparisons, one caution must be explicitly

expressed. Social class differences are con founded with racial difference.

Lower class whites did not learn about politics at a different rate than lower class blacks, although the former were more knowledgeable to begin with. The source of the "gap" was found primarily in the amount of the learning done by middle-class whites, for whom no sizable black counterpart existed for comparison pruposes.

For each type of cognitive information assessed, watching a political series on public television created a substantial growth in political know

ledge. This was so whether one was asking if the youngsters had learned

about the political process in general, about current legislative activities

in particular, or about who the main persons in legislative activities are.

Note an Initial Experimental-Control Group Differences :

On several dependent variables, the experimental and control groups were

not equivalent at the outset of the study. Considering experimental condition as an ordinal variable (non-viewers = 0, viewers = 1), there was a slight

correlation between treatment and initial scores on five variables listed

[ocr errors]

below. A partialling technique was used to eliminate this initial relation

ship from the post-viewing analyses. Partial correlations were computed between experimental treatment and the T2 scores on each variable, controlling

for the ti correlation. The initial associations were also partialled out of

[merged small][merged small][merged small][ocr errors][ocr errors][ocr errors][ocr errors][merged small][ocr errors][ocr errors][ocr errors][merged small][ocr errors][ocr errors][ocr errors][merged small][ocr errors][ocr errors][merged small][ocr errors][merged small][ocr errors][ocr errors][ocr errors][merged small][ocr errors][ocr errors][merged small]

at T2 become marginal when Tl scores are taken into account; the post-viewing association with treatment drops from +. 12 to +.06. These adjusted figures indicate that the programs had little impact on the efficacy variable.

Basic Political knowledge. The slight Tl relationship does not reduce

the T2 correlation of +.27.

Thus, the large post-viewing difference in basic

knowledge appears to be due mainly to the experimental treatment.

National News Watching. Controlling for the original association, the null T2 correlation rises to +.06. However, the adjusted T3 correlation is still negligibly negative. It is clear that the program had no influence over television news viewing, even when the greater initial viewing in the

control group is considered.

Talking with Friends.

The originally positive relationship slightly

attenuates the raw T2 and T3 associations, but each remains significant.

[ocr errors]

Reading Legislature News. Again, the T2 and T3 correlations with ex

perimental treatment are slightly reduced.

The delayed measure still shows

a significant experimental-control group difference.

Interactions Between Treatment and Subject Characteristics

This section examines the interactions between the experimental treatment and certain antecedent variables, including demographics, academic ability, mass media exposure, interpersonal communication, and political orientations. These data are described with correlation coefficients computed be

tween treatment condition (non-viewers = 0, viewers = 1) and the dependent variables, at each of two levels of an antecedent characteristic. The two levels were created from either nominal classifications, e.g., male vs. female, or by dividing the antecedent variable at the median, e.g., knowledgeable vs. less knowledgeable. This procedure provides a simple descriptive

statistic which assesses the relative magnitude of the conditional effect of

classroom viewing on various dependent measures at both the second and third

testing sessions. The first column in Table 5a gives the correlation coeffi

cients across the overall sample. This statistic parallels the ANOVA main

[blocks in formation]

Social Class: working = laborer, operative, service, craftsman, forman;

middle = sales, clerical, professional, technical,

official, managerial

« 이전계속 »