Table 1.
Measures
| Measure | Metric | Modifications | Reliability Cronbach’s coefficient alpha (missing values, %) |
|---|---|---|---|
| HIV Disclosure Matrix (Appendix 1) |
Modified Bauman disclosure matrix (Pequegnat, Bauman, Bray, DiClemente, Dilorio, Hopper et al., 2001) | Added potential targets and reactions; 10 “predetermined” target relationships with 11th target open-ended for “spontaneous” reports | |
| HIV Disclosure Beliefsa Higher scores indicate more positive beliefs |
Berger HIV Stigma disclosure subscale (Berger, Ferrans, & Lashley, 2001) | Six of ten questions from the disclosure subscale were used; two additional questions were asked regarding disclosure to children | Disclosure subscale=0.58 (0) |
| Social Stigma Higher scores indicate lower levels of stigma |
Berger HIV Stigma personalized and public stigma subscales (Berger et al., 2001) | 12 questions were selected; 8 contributed to the personalized subscale and 9 to the public subscale (5 contributed to both) | Personalized stigma subscale=0.72 Public stigma subscale=0.79 All three Berger HIV subscales=0.81 (7, 0.002%) |
| Self-Esteem Higher scores indicate higher self-esteem |
UCSF CAPS HIV Counseling and Testing Self-esteem scale (Coates, Grinstead, Gregorich, Heilbron, & al., 2000) | CAPS self-esteem scale=0.76 (0) | |
| Depression Higher scores indicate worse depressive symptoms |
UCSF CAPS HIV Counseling and Testing Depression scale (Coates et al., 2000) Shona Symptom Questionnaire (SSQ14) (developed and validated in Zimbabwe)b (Nhiwatiwa, Patel, & Acuda, 1998; Patel, Simunyu, Gwanzura, Lewis, & Mann, 1997; Patel, Todd, Winston, Gwanzura, Simunyu, Acuda et al., 1998; Stranix-Chibanda, Chibanda, Chingono, Montgomery, Wells, Maldonado et al., 2005) |
Questions about somatic symptoms were removed from analysis to avoid conflating HIV disease symptoms with depressive symptoms (Kalichman, Rompa, & Cage, 2000) | CAPS depression scale=0.75 (0) SSQ14=0.86 (0) Both depression scales=0.80 (0) |
| Quality of Life (QOL)a Higher scores indicate higher QOL |
Medical Outcomes Study-HIV QOL (Taylor, Dolezal, Tross, & Homes, In press; Wu, Revicki, Jacobson, & Malitz, 1997) | Subscales transformed on 0–100 scale to have mean of 50 and standard deviation of 10(Revicki, Sorensen, & Wu, 1998) Factor analysis used to construct Physical and Mental Health Summary scores (Revicki et al., 1998) |
Range from mental health subscale=0.62 to health distress subscale=0.92 (6, 0.001%) |
a
Missing values imputed with subscale average when more than half the values were present.
b
A second scale was used for the purpose of further validation of the SSQ14; nonetheless, SSQ14 was the preferred depression measurement tool because it was developed locally.