Table 2: Multinomial Logistic Regression for Model 1.

 

Negative

Positive

 

B

SE

Wald

p.

Exp (B)

B

SE

Wald

p.

Exp (B)

Intercept

-0.17

0.15

1.26

0.262

 

-2.44

0.35

47.61

0.000

 

Anal sex (No)

0.23

0.14

2.90

0.089

1.26

-0.11

0.31

0.12

0.728

0.90

Transactional sex (No)

0.15

0.06

5.70

0.017

1.16

0.11

0.16

0.52

0.471

1.12

Alcohol before sex (No)

0.24

0.05

25.53

0.000

1.27

-0.80

0.10

68.22

0.000

0.45

Condom use (Used)

0.44

0.06

65.55

0.000

1.56

1.66

0.10

255.37

0.000

5.24

Number of partners (None)

-0.19

0.08

5.64

0.018

0.82

0.08

0.22

0.14

0.712

1.08

Number of partners (One)

0.21

0.05

16.77

0.000

1.23

0.47

0.14

10.80

0.001

1.59

Union (Polygamous)

0.68

0.06

138.31

0.000

1.97

0.46

0.15

9.07

0.003

1.59

Union (Non-polygamous)

0.71

0.04

319.35

0.000

2.023

0.56

0.10

31.42

0.000

1.76

Note: p < 0.05 is significant. The reference category for the outcome is unknown HIV status. Reference categories for the independent variables are those not in parentheses. Final Model Fitting Chi-square = 784.97, p = 0.000; Pearson χ2 (212, N = 20,848) = 494.431, p < .05; Deviance χ2 (212, N = 20,848) = 440.596, p < 0.05; Cox and Snell = 0.037; Nagelkerke pseudo R2 = 0.049; McFadden = 0.027.