TABLE 8. Regression Analysis of Effect Size on Crisis Intervention Treatment Type and Research Methods Controls
|
Model A regression
|
Model B regression
|
Name of variable
|
Coefficient (SE)
|
t value
|
Coefficient (SE)
|
t value
|
|
| Location |
0.062 (0.067) |
0.132 |
–0.149 (0.447) |
–0.333 |
| Year of study |
–0.002 (0.028) |
–0.083 |
0.014 (0.027) |
0.511 |
| Sample size |
0.000 (0.001) |
0.436 |
–0.000 (0.001) |
–0.106 |
| Statistic method |
–0.312 (0.580) |
–0.537 |
–0.109 (0.571) |
–0.191 |
| Experimental design |
0.157 (0.564) |
0.278 |
0.619 (0.497) |
1.249 |
| Quasi-experimental design |
0.048 (0.477) |
0.101 |
0.058 (0.475) |
0.122 |
| Family preservation |
1.10 (0.679) |
1.624 |
1.068* (0.514) |
2.079 |
| Crisis debriefing |
0.311 (0.425) |
0.635 |
— |
— |
| Multisession crisis intervention or CISM |
1.381 (0.894) |
1.545 |
— |
— |
| Follow-up period |
0.003 (0.036) |
0.082 |
0.036 (0.031) |
1.174 |
Adjusted R2
|
0.17
|
|
0.14
|
|
|
Note. Interpretation of significant effect: family preservation treatment increases the average effect size by 1.1 standard deviation units. SE = standard error; CISM = critical incident stress management.