| Summary: | Background: Research on message effects has focused mainly on text and less attention has been given to text effects in conjunction with other modalities. Text is processed sequentially and undergoes more cognitive control compared to visuals, which are processed almost instantaneously and undergo little cognitive control (Geise, 2017). The EPPM (Witte, 1991) provides theoretical perspectives from which to assess if differing modalities would augment certain message types. It is hypothesized that threat messages would perform better in the visual modality (H1a and H1b) whereas efficacy messages would perform better in the text modality (H2a and H2b). Seasonal influenza was chosen as the context for this study as vaccination rates remain lower than desired locally (“Adults here not keen”, 2017).
Objective: This study aims to examine if these proposed effects would surface and can be used to augment presentation of EPPM message components to maximise the effects of these two message types.
Method: A two-by-two between-subjects experiment with a pre- and post-survey design was used to examine message-modality effects. One hundred and eighty-two Singaporean university students were recruited and randomly assigned to experimental conditions at post-survey. Message content and materials were developed and pretested with 62 participants to reduce the influence of confounding factors like emotion and ensure comprehension. Due to safety restrictions imposed to prevent the spread of COVID-19 at the time of study execution, the planned on-site procedures had to be converted to online surveys.
Analyses: Two-way ANCOVAs were used to test for the hypotheses pertaining to perceived severity (H1a), perceived susceptibility (H1b), and perceived response efficacy (H2a), and intention to vaccinate. Confounders identified from analyses of the baseline data and variables pertaining to COVID-19 were included as covariates. A set of four linear regressions were conducted to test the hypothesis for perceived self-efficacy (H2b) as the data failed the assumption of homogeneity of regression slopes.
Results: No statistically significant main effects of threat message modality were found for perceived severity and perceived susceptibility. Similarly, no statistically significant main effects of efficacy message modality on perceived response efficacy or perceived self-efficacy were found. However, a statistically significant two-way interaction was found for message modality on intention to vaccinate.
Conclusion: The study findings did not support H1 or H2. However, the findings suggest that message-modality combinations where both visual and text is used could be more effective at improving intentions to vaccinate compared to combinations where only a single modality is used. Theoretical implications, practical implications, and future research directions are discussed.
|
|---|