Reexamining the Link Between Cultivation Factors and Viewer Involvement: Investigating Viewing Amount as a Catalyst for the Transportation Process

Abstract

In an effort to expand on prior cultivation research involving narrative processing (Bilandzic & Busselle, 2008), this study examined the interrelationships between viewing amount and the transportation process. Results showed that transportation mediated the relationship between genre-specific television viewing and perceived realism, while overall television viewing directly predicted reduced counterarguing. In addition, structural equation modeling was employed to test a combined cultivation-transportation model leading to changes in viewer excessive drinking perceptions. Findings indicated that genre-specific television viewing indirectly influenced beliefs through increased transportation. Conversely, a significant, direct path was found between overall viewing and excessive drinking beliefs, supporting the original cultivation perspective.

Keywords:

• Cultivation
• Entertainment
• Excessive Drinking
• Transportation

Notes

^** p ≤ .01. ^† p < .10.

For the reliability analyses of the measures, the researcher assessed whether scales were substantially improved by removing specific items and whether the item-total correlation was low. Where no mention of this is given, the reader may assume that the scale reliability would not have been improved by removing any items.

Based on the results of a confirmatory factor analysis involving five dimensions of character involvement—identification, wishful identification, liking, similarity, and parasocial interaction (Moyer-Guse, 2008)—four items from the original identification scale had to be dropped.

Post hoc indirect effect analyses involving only genre-specific viewing were also examined using Hayes' (2012) macro. In all six multiple mediator models, transportation carried the effect of genre-specific exposure on excessive drinking beliefs, bypassing the mechanisms of the transportation process (as evidenced by the bootstrapped confidence intervals not containing zero). However, two significant multiple mediator effects involving identification as Mediator 2 were also found. Thus, in these two analyses, the indirect effect of genre-specific viewing passed through transportation and identification.