Table 3.
Multivariable logistic regression results
| Independent variable | Adjusted odds ratio (99% Confidence Interval) [P Value] | |
|---|---|---|
| Model 1: Use Social Media for Vaccine Information | Model 2: Frequency of Using Social Media for Reading News | |
| Age | ||
| 18–24 | 1 [Reference] | 1 [Reference] |
| 25–34 | 0.79 (0.46, 1.36) [0.270] | 0.87 (0.50, 1.49) [0.495] |
| 35–44 | 0.61 (0.32, 1.14) [0.041] | 0.62 (0.33, 1.16) [0.048] |
| 45–54 | 0.91 (0.44, 1.89) [0.748] | 0.81 (0.39, 1.68) [0.45] |
| 55–64 | 0.54 (0.22, 1.33) [0.078] | 0.44 (0.18, 1.08) [0.018] |
| 65+ | 0.75 (0.30, 1.91) [0.433] | 0.53 (0.21, 1.36) [0.085] |
| Race | ||
| White | 1 [Reference] | 1 [Reference] |
| African American | 1.99 (1.12, 3.52) [0.002] | 2.15 (1.21, 3.82) [0.001] |
| Asian | 0.46 (0.12, 1.74) [0.136] | 0.47 (0.12, 1.79) [0.147] |
| Hispanic | 1.41 (0.75, 2.66) [0.162] | 1.56 (0.82, 2.95) [0.073] |
| Other/unknown | 2.41 (0.51, 11.41) [0.144] | 2.77 (0.58, 13.16) [0.092] |
| Gender | ||
| Female | 1 [Reference] | 1 [Reference] |
| Male | 0.87 (0.55, 1.40) [0.461] | 0.87 (0.54 1.39) [0.447] |
| Other | 0.64 (0.15, 2.66) [0.415] | 0.6 (0.14, 2.61) [0.375] |
| Highest education | ||
| College degree | 1 [Reference] | 1 [Reference] |
| High school degree | 1.72 (1.11, 2.65) [0.001] | 1.7 (1.10, 2.63) [0.002] |
| Master’s or above degree | 0.43 (0.20, 0.90) [0.003] | 0.42 (0.20, 0.89) [0.003] |
| Use social media for vaccine informationa | ||
| Not use social media | 1 [Reference] | |
| Use social media with trusted sources | 1.16 (0.69, 1.93) [0.458] | |
| Use social media alone or with other non-trusted sources | 2 (1.15, 3.46) [0.001] | |
| Using social media for reading news | ||
| Rarely/never | 1 [Reference] | |
| Always | 0.35 (0.20, 0.63) [<0.001] | |
| Sometimes | 0.65 (0.40, 1.07) [0.027] | |
Model 1 examined the association between vaccine hesitancy and social media use for vaccine information while controlling the selected variables constanta; Model 2 examined the association between vaccine hesitancy and the frequency of using social media for reading news while controlling the selected variables constanta (N = 1050)
Predictor Variables used in the regression models include Age, Race, Gender, and Highest Education
aVariables were regrouped into 3 categories (do not use social media, use social media with trusted sources, use social media alone or with other non-trusted sources) to better understand how social media use influences vaccine hesitancy