Table 3.
Log Odds from Logistic Regression Models Assessing Interactions between Retirement, Age, and Partner Betweenness with Respect to Erectile Dysfunction (N = 678)a
| Predictor | Model 1 | Model 2 | Model 3 |
|---|---|---|---|
| Partner betweenness | 1.916** (.389) | 3.227*** (.894) | 1.968** (.402) |
| Proportion kin | .785 (.240) | .776 (.234) | 1.681 (.648) |
| Age (ref: 57–64): | |||
| 65 –74 | 2.284** (.565) | 3.160*** (.863) | 5.106*** (2.189) |
| 75 – 85 | 2.229* (.755) | 2.798* (1.079) | 5.970** (3.236) |
| Partner betweenness x age (65–74) | .334* (.141) | -- | |
| Partner betweenness x age (75–85) | .469 (.302) | -- | |
| Proportion kin x age (65–74) | -- | -- | .240* (.132) |
| Proportion kin x age (75–85) | -- | -- | .162* (.131) |
| F (d.f.) | 4.47*** (31,21) | 4.08** (33,19) | 3.88** (33,19) |
| Fleiss, Williams, Dubro R2 | .159 | .164 | .167 |
*
p < .05,
**
p < .01,
***
p < .001 (two-tailed tests)
a
Estimates are weighted to adjust for probability of inclusion in the analysis and differential non-response. All models are survey-adjusted, and control for all variables included in Table 2.