Table 3. Odds ratios (and 95% confidence intervals) from logistic regression analyses identifying associations between selected characteristics and young women's reporting of coerced sex at Wave 1.
| Characteristic | Bivariate | Total effect | Direct effect |
|---|---|---|---|
| BLOCK 1 | |||
| Town HIV prevalence | |||
| Low (Juapong)(ref) | 1.00 | 1.00 | 1.00 |
| High (Agormanya) | 0.84 (0.57–1.23) | 0.70 (0.46–1.06) | 0.44 (0.27–0.72)** |
| Age | 1.45 (1.33–1.59)*** | 1.46 (1.34–1.60)*** | 1.16 (1.03–1.31)* |
| Household composition | |||
| Lives with both biological parents (ref) | 1.00 | 1.00 | 1.00 |
| Lives with one biological parent | 1.22 (0.72–2.08) | 1.21 (0.68–2.16) | 1.22 (0.65–2.29) |
| Lives with extended family adults only | 1.08 (0.62–1.86) | 1.29 (0.71–2.32) | 1.30 (0.69–2.46) |
| Lives with no family adult | 1.41 (0.76–2.62) | 1.10 (0.56–2.15) | 1.02 (0.50–2.11) |
| Other | 2.13 (0.39–11.52) | 2.57 (0.43–15.27) | 2.94 (0.43–19.97) |
| Household wealth | 0.87 (0.72–1.05) | 0.90 (0.73–1.11) | 0.92 (0.73–1.16) |
| BLOCK 2 | |||
| Behavioral control | 0.62 (0.52–0.74)*** | 0.76 (0.59–0.97)* | 0.85 (0.65–1.10) |
| Relationship quality | 0.77 (0.65–0.91)** | 1.02 (0.80–1.28) | 1.05 (0.82–1.33) |
| Financial support | 0.74 (0.63–0.87)*** | 1.08 (0.87–1.33) | 1.10 (0.88–1.37) |
| Conflict | 1.76 (1.47–2.10)*** | 1.63 (1.34–2.00)*** | 1.61 (1.31–1.98)*** |
| BLOCK 3 | |||
| School enrollment | |||
| No (ref) | 1.00 | 1.00 | 1.00 |
| Yes | 0.25 (0.16–0.37)*** | 0.63 (0.38–1.03) | 0.82 (0.48–1.38) |
| BLOCK 4 | |||
| Ever had boyfriend | |||
| No (ref) | 1.00 | 1.00 | 1.00 |
| Yes | 8.17 (5.34–12.49)*** | 4.51 (2.52–8.07)*** | 4.51 (2.52–8.07)*** |
*
p<.05.
**
p<.01.
***
p<.001.
Notes: Block 1 consists of demographic variables and household composition; Block 2 comprises family process variables; and Blocks 3 and 4 each consist of a single variable, school enrollment status or relationship experience, respectively. Figures in the “bivariate” column are unadjusted odds ratios; those in the “total effect” column are adjusted for all variables in the same and previous blocks; and those in the “direct effect” column are adjusted for all independent variables. ref=reference group.