In the next stage of our analysis we estimate the effect of both structural and
individual measures on adolescent’s number of suicide attempts. Our results indicate that
depression, exposure to violence and exposure to suicidal behaviors are significantly
related with a high suicide attempt event rate, net of neighborhood-level factors.
Final models reveal that even after including both religion variables as predictors
of depression, the only significant interaction remains between depression and
neighborhood level of religiosity. To better understand this cross level interaction we
depict the relationships between neighborhood level religiosity, depression and
adolescent’s probability of suicide attempt in Graph 1. As may be observed in the graph,
adolescents who reside in religious neighborhoods and experience high levels of
depression have a probability of almost 11% to commit one or more suicide attempt.
Adolescents who live in secular neighborhoods and experience the same level of
depression, however, have a probability of 86.6% to commit one or more suicide
attempts. Thus, these results support our hypothesis, suggesting that religious context
reduces adolescent suicide attempts by buffering individual negative emotions and
absorbing its potential to stimulate adolescents’ suicidal behaviors.
With regard to religion, our findings emphasize the importance of future
incorporation of macro and micro explanations into a grand theory of suicide. While we
find that high proportion of conservative population within tract significantly reduces
adolescents’ suicide attempts log event rate, this effect in the equation diminishes when
allowing depression to vary between neighborhoods. At the same time, we find evidence
for a cross-level interaction between individuals’ depression and neighborhood level of
religiosity. We assume that since our new construct of neighborhood level of religiosity
captures regulative religious functions, its effect on depression is significant. This finding
Convention