TPC-Journal-V6-Issue1

The Professional Counselor /Volume 6, Issue 1 68 Table 3. Regression of Self-Injurious Behaviors Variables B Expected B r Aggression 1.220 3.388 .141* Depression 2.381 10.819 .319* Late Adolescence (age) -.234 .791 --- Male -.543 .581 -.090* Caucasian .056 1.057 --- Note: *Significant at p < .001, Nagelkerke R 2 = .111 YRBS 2011 Analysis For the 2011 sample population, 1,300 participants indicated engaging in physical fights four or more times in a year, resulting in the need for medical attention more than once, which fit the criteria for self-injurious aggression (approximately 8% of those surveyed; self-injurious aggression variable). Of those included in analysis, 201 participants indicated that they had attempted suicide four or more times, attempts that did not require medical attention (NSSI variable). Of those students responding, over 4,000 (approximately 29%) indicated feeling sad or hopeless every day for 2 weeks or more in a row during the past year. Feeling sad or hopeless had a weak negative correlation with the NSSI variable with r = -.146, p < .001. Similarly, feeling sad or hopeless had a weak negative correlation with self-injurious aggressive behaviors with r = -.097, p < .001. NSSI and self-injurious aggression had a significant weak positive correlation with r = .195, p < .001. Of those responding to the 2011 YRBS, 7,574 indicated they were Caucasian and 1,629 indicated they were younger than 15 years old. The binary regressive model for the 2011 data indicates a resultant X 2 (4) = 370.27, p < .001. The Nagelkerke R 2 = .241 indicates that this model would only account for approximately 24% of the variance in predicting self-injurious behaviors as defined by items 27 and 28 of the YRBS. Of those surveyed, 69.3% were included in analysis. The Hosmer and Lemeshow test was not significant (χ 2 = 2.39, p = .935), indicating that the predicted probabilities match the observed probabilities. Wald statistics are significant at p < .001 for the item indicating possible depression, age and the variable assessing possible aggressive self-injury (engaging in numerous physical fights). Wald statistics for race were approaching significance at p = .089; however, age and gender were not significant. Therefore, these demographic variables were likely not contributing significantly to the prediction of NSSI as defined in this study. Of those participants who identified as possibly engaging in non-suicidal self-injurious behaviors, 98.5% of cases were correctly classified by the model. The classification of cases was not changed when the variables of non-suicidal aggression, depression, age, gender and race were included. The calculated r statistic for non-suicidal aggression was .30, and .24 for the depression variable, indicating that both likely accounted for 54% of the predictive power of the model. The demographic variables could not be calculated due to their low contribution to the predictive model. While z 2 was significant for age, the Wald statistic itself was not large enough to calculate a standard analogue of r . It is important to note that the lower end of the confidence interval for all variables included in the model was less than one, with the exception of the item variable measuring depressive symptoms.