The Attentional Mechanism of Message Sensation Value: Interaction between Message Sensation Value and Argument Quality on Message Effectiveness

Abstract

Using a secondary data analysis on adolescents’ evaluations of 60 antimarijuana public service announcements, this study examined the role of message sensation value (MSV) as an attention distractor. The results supported the prediction based on the Elaboration Likelihood Model that MSV might be a distractor of attention to reduce ad persuasiveness when the argument quality was high and to facilitate ad persuasiveness when the argument quality was low. Furthermore, this interaction was evident only for adolescents with greater risk for marijuana use, suggesting that high MSV messages were especially distracting for the high risk adolescents. Specific MSV subcomponents contributing to this interaction were explored. Possible explanations for the interaction effect as well as implications for antidrug ad design were discussed.

Keywords:

• Message Sensation Value
• Antimarijuana Public Service Announcements
• Message Effectiveness

Acknowledgments

The research reported here was supported by a grant from the National Institute on Drug Abuse (# DA 12356-02) to M. Fishbein, J. Cappella, and R. Hornik. The views expressed are those of the authors alone.

Notes

1. Southwell (2002) argued that “faces on screen” was a good measure of long term ad recognition because it tapped into “personalization” or “acting out,” the latter being one characteristic of standard MSV coding. Faces on screen also play a primary role in a new message feature measure called “Information Introduced,” which is shown to affect attention to ads (Lang, Bradley, Park, Shin, & Chung, 2006).

2. Similar items measuring perceived ad effectiveness have been used in a previous study with high validity and reliability (Fishbein et al., 2002). Although this scale used some overlapping items as the argument quality scale, they were not significantly correlated (r = .13, p=.31).

3. Additional information on the procedures followed in obtaining and evaluating the effectiveness measures (including the coding procedure for the thought listing data) can be found in Barrett, Ahern, Cappella, Fishbein, and Yzer (2004; a copy of this paper is available from the second author).

4. Although prior literature does not generally find a high correlation between ad liking and the other two message effectiveness measures at the individual level, this result shows that ad liking can be highly correlated with perceived message effectiveness and thought listing at the aggregate level.

5. Although the relatively low correlation between Intense Image subscale and MSV total score did not suggest that the interaction between MSV and argument quality could be carried by Intense Image, this moderate correlation may result from a skewed distribution of Intense Image.

6. Figures for the interaction between these MSV subcomponents and argument quality on message effectiveness are available from the first author upon request.

7. Detailed information about the interaction effect between subcomponents of MSV and argument quality for high risk adolescents is available from the first author upon request.

8. A plausible hypothesis is that high risk adolescents would have more thoughts about ad features in the high MSV condition than low risk adolescents. Thoughts about ad features (e.g., lighting, music, edits, cuts, etc.) and ad content (e.g., ad message and issues brought up in the ad, etc.) were coded separately. Our analyses with MSV, argument quality, and risk as independent variables, and content thoughts or feature thoughts as the dependent measure showed no significant effect of MSV and argument quality. Nor were there significant two- or three-way interactions among the independent measures. There was only a significant main effect for risk on feature thoughts, F(1, 119) = 5.02, p<.03. On average, high risk adolescents reported more thoughts about ad features (M=1.88, SD=2.49) than low risk adolescents (M=1.03, SD=1.35). This difference is consistent with the assumption that high risk adolescents (who are also high sensation seekers) are more attentive to the ad's features than are low risk adolescents.

9. There was a significant interaction between MSV and argument quality for positive thoughts, F(1, 56) = 11.95, p=.001, partial η²=.17. For low MSV ads, strong arguments were associated with more positive thoughts than weak arguments (M _strong=40.67, SD _strong=8.42 vs. M _weak= 33.57, SD _weak=7.02). However, for high MSV ads, strong arguments were associated with fewer positive thoughts (M _strong=38.07, SD _strong=6.16 vs. M _weak= 43.38, SD _weak=6.30). The interaction between MSV and argument quality was not significant for negative thoughts. However, the pattern was opposite to that for positive thoughts. Weak arguments were associated with more negative thoughts for low MSV ads (M _weak=19.00, SD _weak=8.59 vs. M _strong= 13.93, SD _strong=7.19), but fewer negative thoughts for high MSV ads (M _weak=13.44, SD _weak=6.81 vs. M _strong=13.73, SD _strong=7.25). Thus, high MSV does seem to disrupt the predominant thought generated by the ads.

Additional information

Notes on contributors

Yahui Kang

Yahui Kang (MA, University of Connecticut) is a doctoral candidate at the Annenberg School for Communication at the University of Pennsylvania

Joseph Cappella

Joseph N. Cappella (Ph.D., Michigan State University) is Professor of Communication

Martin Fishbein

Martin Fishbein (Ph.D., University of California at Los Angeles) is Professor of Communication