A Methodological Critique of the Consequences Oftelevision News Programs’on the Public

A Methodological Critique of the Consequences

ofTelevision News Programs’on the Public

The 1982 individual-level based longitudinal experiments conducted by ShantoIyengar, Mark D. Peters from Yale University, and Donald R. Kinder from the University of Michigan were designed to test the consequences of television News Programs in public.Iyengar et al. conceptualized “consequences” as in public opinion, or the effect of “agenda setting”, which was indicated by people’s different answers to the same questions in pretest and posttest questionnaires.Thehypothesis held that television news programs significantlyaffected the problems which viewers take to be important; and problems predominantly broadcasted in the evening news furtherinfluenced the viewers’ evaluation of their president. Data was gathered primarily through questionnaires andprocessed using OSL regression in order to draw comparisons cross differentexperimental conditions.In the end, researchers claimed their hypothesis being confirmed.

This article is a critical review of sampling techniques and survey methodology of this study, in which at least five serious errors are revealed (three of them are in the sample procedure and two in data collection process).The presence of even one of these flaws would be sufficient to cast serious doubts on the legitimacy of its results. In combination, they make the data virtually meaningless.

Insufficient Sample Procedure

Sample of this study is highlyunrepresentative. Iyengar et al.alleged that their sample represented “the public” and repeatedly used their data to make generalizations about the entireUS population. However, with the adopted procedure, some people were systematically excluded from the sample while others were over-represented.Problems wereviewedmainly in three aspects: the limited sampling area, the inadequate sample size, and the possibility of biased sample.

First, the initial sampling boundary was confined within only one city. By sampling...