Experimental design can unobtrusively manipulate news coverage to test agenda setting.

Evidence

On the first day, participants completed a questionnaire that covered a wide range of political topics, including the importance of various national problems. Over the next four days participants viewed what were represented to be videotape recordings of the preceding evening’s network newscast. Unknown to the participants, portions of the newscasts had been altered to provide sustained coverage of a certain national problem. On the final day of the experiment (24 hours after the last broadcast), participants completed a second questionnaire that again included the measures of problem importance. — John T. Jost and Jim Sidanius, Political Psychology, 1452

Categorization

• XREF: Bobrow’s point that experiments can establish causation where dynamic analysis can’t; echoes the existing methodological concern with causal inference.