Table 4.
Multivariate regression results for SHS exposure, smoke-free policy noncompliance, and presence of restrictive home smoking rules.
| Explanatory Variable | Smells Smoke Anywhere in Development | Smoke Entering Unit | Respondent Smoking In-Unit (Restricted to Smokers) | Someone Else Smoking In-Unit | Restrictive Home Smoking Rules | Number of Places Smoke Smelled in Development |
|---|---|---|---|---|---|---|
| N = 939 | N = 1002 | N = 144 | N = 992 | N = 1002 | N = 939 | |
| Odds Ratio | Coefficient (95% Confidence Interval) | |||||
| (95% Confidence Interval) | ||||||
| Wave (ref = Wave 1) | ||||||
| Wave 2 | 0.23 ** (0.13–0.44) | 0.39 ** (0.23–0.67) | 0.01 **(0.00–0.16) | 0.45 * (0.23–0.88) | 3.63 ** (2.20–6.00) | −0.70 ** (−0.93–−0.47) |
| Repeat response (ref = non-repeat respondents) | ||||||
| Repeat respondents | 1.51 (0.80 –2.88) | 2.53 * (1.22–5.24) | 9.97 (0.33–298.78) | 1.17 (0.58–2.38) | 1.08 (0.68–1.70) | 0.18 (−0.13–0.49) |
| Age (ref = 18–64) | ||||||
| Aged 65+ | 0.48 * (0.27–0.87) | 0.22 ** (0.10–0.47) | 1.56 (0.05–50.04) | 0.49 (0.24–1.01) | 1.60 * (1.03–2.47) | −0.59 ** (−0.88–0.30) |
| Sex (ref = female) | ||||||
| Male | 0.81 (0.43–1.55) | 0.33 ** (0.16–0.69) | 4.14 (0.19–91.73) | 1.25 (0.63–2.51) | 1.78 * (1.08–2.94) | −0.17 (−0.48−0.14) |
| Race (ref = White) | ||||||
| Black or African American | 0.80 (0.36–1.78) | 1.01 (0.44–2.33) | 0.21 (0.002–21.98) | 0.88 (0.39–1.99) | 0.80 (0.45–1.42) | 0.10 (−0.30–0.50) |
| Asian | 0.50 (0.14–1.74) | 0.54 (0.14–2.15) | 0.00 (0.000–570.31) | 0.51 (0.10–2.60) | 1.40 (0.52–3.77) | 0.04 (−0.59–0.66) |
| Native Hawaiian, Pacific Islander, American Indian, or Alaska Native | 1.62 (0.23–11.24) | 8.97 * (1.66–48.63) | 0.01 (0.000–12.20) | 1.83 (0.35–9.52) | 0.41 (0.13–1.30) | 0.38 (−0.49–1.26) |
| Multi-race | 0.65 (0.15–2.92) | 1.81 (0.34–9.76) | 1.95 (0.34–11.32) | 0.82 (0.25–2.71) | 0.43 (−0.41–1.27) | |
| Ethnicity (ref = not Hispanic/Latino) | ||||||
| Hispanic or Latino | 1.03 (0.47–2.24) | 0.69 (0.30–1.63) | 0.90 (0.005–168.36) | 0.29 ** (0.12–0.70) | 0.93 (0.53–1.65) | 0.28 (−0.11−0.67) |
| Smoking status ref = non-current smoker) | ||||||
| Current smoker | 1.12 (0.48–2.65) | 0.32 * (0.13–0.80) | 11.01 ** (3.98–30.45) | 0.08 ** (0.03–0.18) | -0.41* (-0.79–−0.04) | |
| Constant | 37.92 ** (9.35–153.73) | 17.58 ** (5.05–61.17) | 251.72 (0.98–64,545.99) | 0.04 ** (0.009–0.16) | 2.00 * (1.03–3.91) | 3.37 ** (2.90–3.83) |
Note: Includes only respondents who report living in their unit for at least 1 year. All models accounted for within-unit correlation and controlled for PHA-level fixed effects (not shown in regression table). All models were conducted using logistic regression, except for the model of the number of places respondents smelled smoke within their building, which was conducted as a linear regression. For logistic regression models, we present odds ratios rather than coefficients. Due to cell size limitations, the model of respondents smoking in units did not include an indicator for model specification: outcome = β0 + β1 (Wave 2) + β2 (repeat respondent) + β3 (age 65+) + β4 (male) + β5 (current smoker) + β6-9 (PHA indicators) ** p < 0.01, * p < 0.05. Bolding indicates variable name.