The Effects of Scene Changes and Semantic Relatedness on Attention to Television

Abstract

Two laboratory experiments tested the proposition that attention to television is the product of an interaction between message structure and content. Cuts (structure) were hypothesized to affect attention differently, depending on whether they linked related or unrelated content. Subjects viewed six semantically related and six semantically unrelated sequences of television. In Experiment 1, attention was measured using reaction times to a secondary task at one video frame, 1 second and 2 seconds following a cut. Unrelated sequences required more attention 1 second after the cut. Attention to related sequences decreased over the course of the entire sequence, whereas attention to unrelated sequences remained constant over time. Experiment 2 assessed attention during the first second after the cut. Related sequences produced longer reaction times immediately following the cut at 10 and 20 frames; unrelated sequences produced longer reaction times further from the cut, at 20 frames, and 1 second. This pattern of results is explained using a dual process model of attention to television.