Table 6.
Logistic regression models for self-biting presence (‘self-abuse’ subcategory) 1.
| Independent Variables | Unadjusted Odds Ratio and p-Value | Adjusted Odds Ratio * | 95% CI * | p-Value * | Positively and Significantly Predicts Self-Biting? * (Result the Same if Age, Sex and Scanning Level not Controlled for?) |
|---|---|---|---|---|---|
| Threat-biting presence vs. absence (rarely scanned group) |
11.00 (0.022) | 9.02 | 0.86–94.56 | 0.034 | Yes (Unadjusted relationship is also positive and significant) |
| Threat-biting presence vs. absence (moderately scanned group) |
3.00 (0.167) | 2.76 | 0.29–25.84 | 0.188 | No (Unadjusted relationship is also positive and non-significant) |
| Threat-biting presence vs. absence (frequently scanned group) |
1.00 (0.495) | 0.89 | 0.31–2.53 | 0.587 | No (Unadjusted relationship is also non-significant) |
| Self-hitting presence vs. absence 2 | 6.48 (0.003) | 4.04 | 0.96–16.98 | 0.029 | Yes (Unadjusted relationship is also positive and significant) |
| Self-injurious behaviour presence vs. absence 2 | 7.19 (<0.001) | 5.43 | 2.18–13.49 | <0.001 | Yes (Unadjusted relationship is also positive and significant) |
* Adjusted for sex, age, and scanning level. 1 One-tailed p-values are reported here. Significant p-values for adjusted positive relationships are bolded for emphasis.2 Interaction not tested due to insufficient observations.