Cross sectional study question

#1
I am designing a cross sectional study involving the comparison of 2 samples of adolescents between the ages of 11 and 14. One sample will be composed of adolescents who still wet the bed. The other sample will be controls. I have two questions. I am bad at this class, so these questions might be stupid.

Question #1 Can I strategically recruit my controls so that this sample is similar to the bedwetting sample in terms of representation of low SES participants in order to control for the confounding effects of SES? Or do the rules of cross sectional studies dictate that I need to recruit a my controls randomly?

Question #2 I am trying to investigate the prevalence of delinquency, psychosocial problems, and traumatic life events in the bedwetting sample as it compares to the controls in order to see whether there is a significant difference between the two groups. All three of these variables will be continuous. What are some tests that I can perform to analyze my data? For each variable (delinquency, psychosocial problems, and traumatic life events), should I perform a paired t test followed by logistic regression? It is my understanding that a paired t test would tell me whether there was a significant difference between the control group and the "diseased" group for each predictor and I thought that logistic regression would show me how strong or weak each predictor is in terms of the relationship with bedwetting. Is that right?
 
Last edited: