Help with path analysis

#1
Hi everyone,

I am a clinical psychology student trying to use confirmatory path analysis to determine whether a model using gender (categorical), gender stereotyped self-concept discrepancy, self-esteem, and parental support to "predict" suicidal ideation for a normal self-selected sample of 18 and 19 year old post-high school students. All of the other variables are measured using likert-scaled measures. My problem is that I have only been able to recruit 92 women and 21 men in a year and a half. I am trying to determine whether this discrepancy in the number of male and female participants will be problematic for the analysis, but I am unable to find any resources that have any answers. I have been trying on my own for two weeks since consulting with my committee. I can not move onto data analysis without this answer, which is impacting my timeline for graduation. Any help would be greatly appreciated. Thanks.
 

JohnM

TS Contributor
#2
Yes, if gender is a predictor variable in your model, then 21 men vs 92 women would present a problem - the unbalanced sample sizes may also affect the analysis/interpretation of the other predictors as well (you may see a sample size "effect" rather than a real effect).

There is really nothing else, in my opinion, that can be done, other than finding 71 more men to participate.

Have you tried offering an incentive, such as $5? - I know 5*71 is a lot of money, but it may reap benefits.

Another idea - approach professors at your school and mention your sample size problem, and ask them to offer extra-credit in their class for participating in your study.

How about advertising on campus (flyers)?

If all it takes is answering a survey, I'm sure you could twist a few arms....
 
#3
Thanks JohnM for having a simple answer. I do have an incentive, have posted flyers, placed newspaper ads in campus papers, have had several college professors let me come into their classrooms, and I have sat in student quads asking everyone who walks by. The problem I believe is that I am asking people to answer questions about their thoughts and feelings about their lives and their relationship with their parents. It was the best wording I could come up with to be vague about my dependent variable-suicidal ideation. Seems to be the last thing 18 and 19 year old men want to do. Maybe that's the real evidence of gender role socialization. I'll go back to my committee and see what options they will allow me regarding making any changes to my study at this point. Thanks again.