Understanding how the message appeal of moral beauty influences advertising effectiveness under mortality salience

Abstract

Consumers may be reminded of life mortality by some life events and media content. Mortality salience as a contextual factor has not been sufficiently investigated in the existing literature of advertising. The present research analyzes how the interplay between mortality salience and moral beauty influences consumers’ perceptions of life meaning and their responses to advertisements. Results of two studies confirmed that participants perceived increased life meaning when viewing advertisements that presented moral beauty. In Study 1, participants who were reminded of death evaluated an advertisement that presented moral beauty more favorably than an advertisement without such content. In Study 2, younger (i.e., age 18–25) and older participants (i.e., 55 and above) were recruited and the results indicated that the phenomenon observed in Study 1 was only prominent among younger participants. These results are believed to provide interesting insights to marketing communication both theoretically and practically.

Keywords:

• Message appeal
• moral beauty
• mortality salience
• terror management
• life meaning
• age difference
• advertising effectiveness

Acknowledgements

The author would like to thank the two anonymous reviewers and the deputy editor Professor Kim Sheehan for their invaluable assistance during the review process. The author would also like to thank Rida Ryu for developing the stimuli in Study 1 as well as Nanlan Zhang and Chris Noland for coding participants’ responses.

Notes

1. Before testing the hypotheses, an MANCOVA test was conducted with message appeal and mortality salience manipulation as the independent variables, perceived life meaning and ad attitude as the dependent variables, and product involvement and ad familiarity as covariates. The covariates were added for ensuring that participants’ involvement with insurance services and their familiarity with the ad content would not account for the final results. The results showed that ad familiarity was not a significant covariate, but product involvement was. Therefore, product involvement was added as a covariate in the following analyses.

2. Before testing the hypotheses, an MANCOVA test was conducted with message appeal, mortality salience manipulation, and age as the independent variables, perceived life meaning and ad attitude as the dependent variables, and product involvement and ad familiarity as covariates. The covariates were added for ensuring that participants’ involvement with two product/service categories (i.e., insurance vs. automobile) and their familiarity with the ad content would not account for the final results. The results showed that ad familiarity was not a significant covariate, but product involvement was. Therefore, product involvement was added as a covariate in the following analyses.

3. The mediation analyses were run in SPSS 21.0 with Preacher and Hayes’s macro (2008). The results indicated that the relationship between message appeal and ad attitude was mediated by perceived life meaning in Study 1, as the 95% confidence interval (CI) for the indirect effect did not include zero (ab = −1.07, 95% CI = −1.52 to −0.67). The relationship between message appeal and ad attitude was also mediated by perceived life meaning in Study 2, as the 95% confidence interval (CI) for the indirect effect did not include zero (ab = −1.49, 95% CI = −2.00 to −1.06).

4. A moderated moderation analysis with 5,000 bootstrapped samples was conducted using model 3 of the PROCESS macro for SPSS (Hayes 2012) to test the three-way interaction between mortality salience manipulation, message appeal, and age on ad attitude (product involvement was added as the covariate). The results revealed that the three-way interaction was non-significant, as the 95% confidence interval for the three-way interaction effect included zero (B = −0.05, 95% CI = −0.13 to 0.04).

5. The research assistants coded participants’ writing in the prime task in both studies. Each response was coded for topic relevance (how relevant of the response is to death or seeing a dentist? 1 = not relevant at all, 5 = very much relevant) and number of words. In Study 1, the mean score for topic relevance was 4.69 (SD = 0.65) and it was significantly greater than the middle point of the scale (i.e., 3), t (116) = 28.19, p < .001. The average words for one response was 76.07 (SD = 48.74). In Study 2, the mean score for topic relevance was 4.93 (SD = 0.26) and it was significantly greater than the middle point of the scale (i.e., 3), t (228) = 114.33, p < .001. The average words for one response was 83.53 (SD = 35.61).