| Variable | Gender | Test of proportions or t test | |
| Male (n = 123) | Female (n = 133) | ||
| Demographic characteristics | |||
| Age (% in each category) in years | |||
| <20 | 1.6 | 0.3 | NS |
| 20–29 | 33.3 | 31.6 | NS |
| 30–39 | 37.4 | 31.6 | NS |
| 40–49 | 18.7 | 25.6 | NS |
| 50–59 | 4.1 | 7.5 | NS |
| >59 | 4.9 | 3.0 | NS |
| Marital status (% in each category) | |||
| Married, living with partner | 47.2 | 32.3 | 0.02 |
| Single, widowed, separated/divorced | 52.9 | 67.7 | 0.02 |
| Self-reported race (% in each category) | |||
| African | 65.9 | 71.2 | NS |
| Multiracial | 25.2 | 22.7 | NS |
| Indian | 4.9 | 4.6 | NS |
| White | 4.1 | 1.5 | NS |
| Province (% in each category) | |||
| Eastern Cape | 36.6 | 27.8 | NS |
| Western Cape | 29.3 | 40.6 | 0.06 |
| KwaZulu-Natal | 34.2 | 31.6 | NS |
| No in household (% in each category) | |||
| 1–2 | 22.8 | 13.5 | 0.05 |
| 3–8 | 61.8 | 71.4 | NS |
| >8 | 15.4 | 15.0 | NS |
| Socioeconomic characteristics | |||
| Formal education (% in each category) in years | |||
| <7 | 15.5 | 12.0 | NS |
| 8–12 | 66.7 | 63.2 | NS |
| >12 | 17.9 | 24.8 | NS |
| Income from regular employment (%) | 77.2 | 75.9 | NS |
| Sources of non-employment income (mean no of sources) | 0.58 | 0.82 | 0.03 |
| Income in past 30 days (median; US$)* | 279 | 241 | NS |
| Income in past 30 days, adjusted for household size (median; US$)* | 70.8 | 50.2 | NS |
| Subjective social status | |||
| Community ladder (mean) | 4.50 | 4.35 | NS |
| Country ladder (mean) | 3.93 | 4.01 | NS |
| Creditworthy ladder (mean) | 4.60 | 4.08 | NS |
| Stressful life events | |||
| Events not including birth and catastrophe (mean no of events) | 1.60 | 1.64 | NS |
| Birth in household (%) | 8.9 | 10.5 | NS |
| Catastrophe in household (%) | 7.3 | 5.3 | NS |
| Decision-making (mean no of items) | |||
| Respondent alone | 3.37 | 4.84 | 0.05 |
| Respondent with partner | 6.55 | 6.41 | NS |
| Partner/other alone | 3.08 | 1.75 | 0.01 |
NS, not significant, p>0.10.
*US$1 = 7.3 South African rands. For statistical analysis, the logarithm of this variable was computed because of positive skew.