For each of the continuous covariates (age and financial stress), we ran model specifications including its square term to test whether its relationship with the outcome is nonlinear, and then include the squared term for the final model if a statistically significant association, or a trend toward significance, was found between the square term and the outcome.
← all excerpts
The Gap in Mental Health Service Utilization Among United Methodist Clergy with Anxiety and Depressive Symptoms.
1
—
—