Table 3. Logistic set regression: condom use at most recent anal intercourse (N = 356).
Model | −2 log L | Overall model | Δ −2 log L | |
---|---|---|---|---|
Set 1: cultural set | 399.84 | X2(3) = 2.72 | ||
Set 2: partner set | 377.92 | X2(5) = 24.64* | 21.92** | |
Set 3: physical set | 370.83 | X2(7) = 31.74** | 7.09* | |
Set 4: condom communication | 309.30 | X2(8) = 93.27** | 61.53** | |
| ||||
Final model | Coefficient | Wald X2 | Odds ratio | 95% CI |
| ||||
Set 1: cultural set | ||||
Brazilian | 0.53 | 2.21 | 1.70 | 0.85–3.41 |
Colombian | 0.09 | 0.08 | 1.10 | 0.57–2.13 |
Latino beliefs inventory | −0.17 | 0.52 | 0.85 | 0.53–1.34 |
Set 2: partner set | ||||
Main partner | −0.70 | 4.88* | 0.49 | 0.27–0.92 |
Seroconcordance | −0.68 | 4.35* | 0.51 | 0.27–0.96 |
Set 3: physical set | ||||
Sexual desire | 0.03 | 0.04 | 1.03 | 0.73–1.46 |
Intoxication | −0.65 | 4.07* | 0.52 | 0.28–0.98 |
Set 4: condom communication | 2.23 | 50.83** | 9.31 | 5.04–17.19 |
P < 0.05,
P < 0.0001