Extending the On-line Processing Model: Constructing Attitudes on Personified Issues
Francis K. Neely and Michael Guge, State University of New York at Stony Brook

 Abstract

      Recent models of political attitude construction have assumed a memory-based process dominates, and an on-line process is exceptional (Zaller and Feldman 1992, Zaller 1994). However, studies of how we process information about political candidates show that people are on-line processors (Lodge, McGraw, and Stroh 1989; Lodge, Steenbergen, and Brau 1995). Only one study to date has considered on-line processing of a political issue (McGraw and Pinney 1990). Meanwhile, attaching main characters to news stories--i.e., episodic framing--has been shown to influence how receivers interpret the information, at least regarding assignment of responsibility on political issues (Iyengar 1991). We identify a gap in the literature, and examine the influence of episodic framing of a political issue on the nature of cognitive processing.
     We suspect that on-line processing of issues does occur, although it is not clear when or where. Does presenting an issue through the experience of a protagonist tend to meld the issue with that main character, and therefore, enhance on-line processing? Or do the main character and the issue remain distinct attitude objects, in which case the person in the story may present an obstacle to on-line processing? Through experimental method we test the former expectation--that the issue and the person become one attitude object, promoting on-line processing.
      We find evidence of just the opposite. Overall, personification of our issue increased the memory-judgment correlation (evidence of a memory-based process). In addition, the level of issue salience and general political knowledge are considered. Our data suggest that in the personified frame, only knowledgeable subjects who care little about the issue use an on-line process, and that all others reading the personified frame draw much more on memory to construct their judgments. We present this study as a preliminary look, and recognize the limitations of the findings.