Skip to main content
. 2022 Jan 7;12:819573. doi: 10.3389/fpsyt.2021.819573

Table 2.

Logistic regression model for approval of any type of coercion.

CI
B SE p OR Lower Upper
Endangering behavior
None vs. self-endangering 0.748 0.198 <0.001 2.112 1.433 3.115
None vs. endangering others 0.477 0.207 0.021 1.611 1.074 2.418
Diagnosis
Psychosis vs. alcohol dependency −0.726 0.193 <0.001 0.484 0.332 0.706
Psychosis vs. BPD −0.516 0.187 0.006 0.597 0.413 0.862
Familiarity
Friends vs. none −0.730 0.269 0.007 0.482 0.285 0.816
Family vs. none −0.467 0.256 0.068 0.627 0.379 1.035
Self vs. none −0.386 0.256 0.132 0.680 0.411 1.123
Desire for social distance 0.038 0.026 0.144 1.038 0.987 1.092
Perceived dangerousness 0.137 0.024 <0.001 1.146 1.094 1.201
Treatment useful 1.543 0.330 <0.001 4.680 2.453 8.928
Gender (vignette) 0.169 0.156 0.279 1.184 0.872 1.609
Gender (respondent) 0.189 0.153 0.216 1.208 0.895 1.631
Constant −4.649 0.614 0.000 0.010

B = unstandardized regression weight, SE, standard error; CI, Confidence interval; p, p-value; OR, odds ratio; BPD, borderline personality disorder; vs., versus. R2 = 0.196 (p < 0.001).