Abstract:

Numerous studies have discussed the relationship between self-efficacy (SE) and online health information seeking behavior (OHISB), but the results were divergent. Theoretically, according to the health belief model, there was a significant positive correlation between the two variables, while the comprehensive model of information seeking suggested that there was no correlation between the two variables, and the risk perception attitude framework argued that the relationship between the two variables is correlated in different directions under different levels of risk perception. Empirically, the effect sizes of this relationship reported in the existing empirical research were far from consistent, with r values ranging from -0.10 to 0.71. Therefore, a meta-analysis was conducted in this paper to explore the strength and moderators of the relationship between SE and OHISB.
Through literature retrieval, a total of 46 independent effect sizes (44 studies, 21912 participants) were found within the criteria of the meta-analysis. On the basis of the characteristics of studies, the random-effects model was selected. The heterogeneity test illustrated that there was significant heterogeneity among 46 independent effect sizes, indicating that the random-effects model was appropriate for subsequent meta-analyses. In addition, no significant publication bias was found in the included studies according to the results of Funnel Plot, Classical fail-safe indicator, and Trim and Fill.
The main effect analysis indicated that there was a significant positive correlation between SE and OHISB (r =0.27, 95% CI [0.21，0.33]). The subgroup analysis and meta-regression revealed that the relationship between SE and OHISB was moderated by cultural background and the health status of the participants. The results showed that: (1) There is a positive correlation between SE and OHISB, the higher individuals’ level of SE is, the more willing and more frequently they would seek health information on the internet. (2) From the perspective of cultural background, the correlation between SE and OHISB was stronger under the background of collectivism than individualism. (3) From the characteristics of participants, the correlation between SE and OHISB was higher in the sick group than in the healthy group; Gender did not moderate the relationship between SE and OHISB, the correlation between SE and OHISB may be stable across genders. (4) In terms of measurement indicators, both internet self-efficacy and health self-efficacy were positively correlated with OHISB, as so was SE with the intention, the frequency and duration, and the experience of online health information seeking. The measurement indexes of SE and the measurement indexes of OHISB did not moderate the relationship between SE and OHISB.
This meta-analysis has both theoretical and practical significance. In theory，this study validated that SE is closely related to OHISB, which initially clarified the academic arguments between the health belief model, the comprehensive model of information seeking and the risk perception attitude framework. We found that individuals’ online health information seeking behavior was affected by both media-related and health-related belief factors, the findings confirmed the adaptation of the health belief model theory in cyber society. In practice, we suggest that the governments and health care practitioners could help people improve the level of SE through multiple ways, such as conducting internet skills training and organizing lectures on health education, thus encouraging more people to seek health information on the internet and effectively excavating the potential of internet in improving the ability of personal health management.
Future research would focus on the standardization of measurement tools of SE and OHISB, and add longitudinal and experimental studies to further explore and clarify the causal relationship between SE and OHISB.

Key words: self-efficacy, health information seeking, internet use, meta-analysis