Table 3.
Multiple logistic regression models predicting difference between pre-test and post-test scores on Link’s Devaluation-Discrimination Scale
|
3-1 Results of multiple logistic regression analysis using only demographic variables to predict considerably decreased stigma* | ||||||
|---|---|---|---|---|---|---|
|
Variables |
Coefficients |
Std. error |
P |
Odds ratio |
95% CI |
|
| Lower | Upper | |||||
|
Employment status |
|
|
|
|
|
|
| Full-time |
0.512 |
0.224 |
0.022 |
1.669 |
1.077 |
2.587 |
| Part-time |
0.640 |
0.255 |
0.012 |
1.897 |
1.151 |
3.125 |
| Other** |
|
|
0.032 |
|
|
|
|
Occupation |
|
|
|
|
|
|
| Production labor service** |
|
|
0.004 |
|
|
|
| Transportation and communication |
0.716 |
0.298 |
0.016 |
2.045 |
1.141 |
3.666 |
| Sales and marketing |
0.189 |
0.215 |
0.380 |
1.208 |
0.792 |
1.840 |
| Service industry |
0.442 |
0.237 |
0.062 |
1.557 |
0.978 |
2.478 |
| Professional |
0.366 |
0.195 |
0.061 |
1.442 |
0.984 |
2.115 |
| Homemaker |
0.689 |
0.280 |
0.014 |
1.992 |
1.151 |
3.448 |
| Others |
−0.463 |
0.295 |
0.117 |
0.629 |
0.353 |
1.123 |
|
Constant |
−0.752 |
0.246 |
0.002 |
0.472 |
|
|
|
3-2 Results of multiple logistic regression analysis using the significant demographic variables and four measures at pre-evaluation to predict considerably decreased stigma* | ||||||
|
Variables |
Coefficients |
Std. error |
P |
Odds ratio |
95% CI |
|
| |
|
|
|
|
Lower |
Upper |
|
Employment status |
|
|
|
|
|
|
| Full-time |
0.315 |
0.170 |
0.064 |
1.370 |
0.982 |
1.912 |
| Part-time |
0.564 |
0.231 |
0.015 |
1.758 |
1.117 |
2.767 |
| Other** |
|
|
0.037 |
|
|
|
|
Knowledge score |
−0.150 |
0.061 |
0.015 |
0.861 |
0.763 |
0.971 |
|
Pre-program Link’s Devaluation-Discrimination Scale |
0.359 |
0.024 |
<0.001 |
1.432 |
1.366 |
1.502 |
|
Social distance |
−0.162 |
0.021 |
<0.001 |
0.850 |
0.815 |
0.887 |
| Constant | −8.576 | 0.890 | <0.001 | 0.000 | ||
*Dependent variable: 1 = stigma-decreased, 0 = stigma-increased. The two groups were defined by subtracting the total post-test score on Link’s Devaluation-Discrimination Scale from the total pre-test score on Link’s Devaluation- Discrimination Scale. The significant variables were selected using forward stepwise regression. Pre-evaluation measures include basic knowledge, pre-program Link’s Devaluation-Discrimination Scale score, ability to identify and distinguish schizophrenia from other conditions, and social distance.
**Reference category.