Table 2. Predictors of message-sending using logistic regressions.
| Outcome: Odds of Sending a Message | |||
|---|---|---|---|
| (1) | (2) | (3) | |
| A. Received Public Health Message | 1.12 | 0.78 | 1.2 |
| [0.85,1.49] | [0.52,1.16] | [0.91,1.60] | |
| B. Received More Effective Prevention Technology | 1.06 | 0.92 | 1.03 |
| [0.81,1.38] | [0.63,1.35] | [0.79,1.35] | |
| C. Baseline Illness Rate = 0.30 | Ref. Group | Ref. Group | Ref. Group |
| D. Baseline Illness Rate = 0.50 | 2.91** | 2.11** | 1.75** |
| [2.08,4.09] | [1.28,3.48] | [1.26,2.44] | |
| E. Baseline Illness Rate = 0.70 | 4.61** | 3.82** | 2.01** |
| [3.27,6.48] | [2.33,6.27] | [1.42,2.85] | |
| F. Prevented | 0.83 | 1.47** | |
| [0.43,1.62] | [1.12,1.94] | ||
| G. Fell Sick | 0.62 | 0.88 | |
| [0.27,1.44] | [0.67,1.16] | ||
| H. Prevented X Fell Sick | 2.3 | 1.46* | |
| [0.89,5.96] | [1.06,2.00] | ||
| I. Number of Messages Received in Previous Round | 3.46** | ||
| [3.05,3.92] | |||
| Rounds | All | 1st Only | Rounds 2–15 |
| Mean of Dependent Variable in Reference Group (No Prevention, Not Sick) | 0.28 | 0.20 | 0.18 |
| P-value: G+H = 0 | 0.14 | 0.00 | |
| Number of Observations | 10185 | 679 | 9506 |
Notes: Table shows logistic regression results for predictors of messaging overall (Column 1) after the first round only (Column 2), and after rounds 2–15 (Column 3). All regressions include the following controls: the age, gender, marital status, ethnicity, occupation, parental status, education, income, and prevention behavior of the individual (takes vitamins had a flu shot in the past year, favors child vaccination, number of dentist visits in past year, sunscreen use). Coefficients are expressed in terms of odds ratios and 95% confidence intervals are in brackets. Standard errors are adjusted for clustering by individual.
*p<0.05
**p<0.01