What statistical analysis do I use? (psych research)?

I'm writing a research proposal paper, and I'm not sure what statistical analysis method I should use. The proposal involves investigating whether or not the use of parental harsh discipline leads to the onset of childhood OCD. So far, I basically have that the participants will include 40 pediatric OCD patients and a healthy control group of 40 children. Both the clinical and nonclinical sample group of children will report on the kind of discipline that they have been subjected to by their parents using a self-report questionnaire (or maybe an interview, i haven't decided yet). Anyway, I'm just not sure which statistical analysis to use to compare the data from the group of kids with OCD and the group of healthy kids in the control group to show the difference in the abuse used by the parents of the kids in each group. Any advice would be great, and if you have any other ideas, feel free to add them..i'm an undergrad and eager to learn :) Thanks!!

What statistical analysis do I use? (psych research)?
I agree with the t-test method. The t-test assesses whether the means of two groups are statistically different from each other. This analysis is appropriate whenever you want to compare the means of two groups, and especially appropriate as the analysis for the posttest-only two-group randomized experimental design. But also using some of the aspects of the Two-Group Comparison would work as well.
Reply:If you take a qualitative approach to the research there not a lot of statistics. But you have to define your method first then depending on that you figure out statistical analysis. As long as you do a quantitative research questionnaire or interview with closed answers. You may need a T test to determine there is a difference between the means or/and some type of correlation factor to link a type of discipline to the clinical sample or to the nonclinical.
Reply:Ok so in order to see if the difference between the OCD and non-OCD group is significant you should use a t-test because you only have 2 groups. But depending on how you look at it, you might want to use a 2x2 ANOVA because you have OCD/non-OCD and parent abuse/no abuse.
Reply:I guess my question is not what test to use, but whether you will get any meaningful data from either a self-report questionnaire or an interview. I am double-certified (state dept of children and families) for sexual abuse interviewing, and I know how difficult it is to get usable information from children about their abuse by their parents.





Also, you switch in your question from "harsh discipline" to "discipline" to "abuse." Which is it? And how are you defining each of those? Obviously "harsh" discipline is worse than "discipline," but are you thinking physical, emotional, or what? I would think you'd define each as both what it is and what it is not. I also suggest that you pick one term and use it consistently every time you refer to it.





Also, are you looking for some kind of cause-effect relationship between abuse and OCD? This is not the same as correlation, of course, where there may be a statistical relationship but no real causation. And which are you more interested in finding out about: obsessive thinking, or compulsive actions?





And finally, I wonder on what previous research you base the possibility of ANY relationship between parental discipline or abuse and OCD? Here's a statement from one web site:





"Cause of OCD -- No one is really sure of the exact causes of obsessive compulsive disorder. It is thought that OCD is caused by a defect in the your brain, which prevents adequate communication between nerve cells. In particular, OCD sufferers seem to have a problem with the frontal lobe and the basal ganglia, a structure deep within the brain. OCD sufferers also tend to have low levels of serotonin in their brains, a neurotransmitter responsible for helping parts of your brain communicate with other parts. Other causes of OCD may rest with genetics. No specific genes that cause OCD have yet been discovered, however, it is believed that some cases of OCD may be inherited. In particular, childhood OCD seems to run in families and more than a third of all adults with OCD claim their disorder began in childhood."





Here's a statement from the National Institutes of Health:


"Scientists at the National Institutes of Health's (NIH) National Institute on Alcohol Abuse and Alcoholism (NIAAA) have identified a previously unknown gene variant that doubles an individual's risk for obsessive-compulsive disorder (OCD). The new functional variant, or allele, is a component of the serotonin transporter gene (SERT), site of action for the selective serotonin reuptake inhibitors (SSRIs) that are today's mainstay medications for OCD, other anxiety disorders, and depression." http://www.medicalnewstoday.com/articles...





You say you are investigating "whether or not the use of parental harsh discipline leads to the onset of childhood OCD." I wonder if the investigation should be turned around, to explore the extent to which children with OCD are harshly disciplined or abused because of their OCD...the OCD thus "causing" (or at least related to) the abuse.





Just some thoughts...hope you succeed with this project.