Hi Rawan,

It is difficult to answer this question without telling us what type of relationships you are looking for. For example, are you interested in bivariate correlations only? Are you more interested in assessing a specific structure of relationships through path analysis? Are you looking to see how various groups differ on outcomes? The list goes on, but I think it is important to state exactly what you want to examine. Then we can provide some recommendations.

Cory

"My supervisor would like to include scales and tests" is not a research question. And I'm guessing that your supervisor may be uncomfortable with research methods that cannot be fed into the old t-and-p grinder.

I would go back to your supervisor and ask what constructs in your research question could be measured with scales and tests in a more meaningful way than your approach, which is listening to people. Consider their answer and see if you think that the quantitative measures they suggest would provide a useful way of illuminating your study question from a complementary angle – would they are the qualitative results build a better picture than the qualitative results alone.

If you feel that the suggested quantitative measures are testing a different research question, speak up. An undergraduate project should have one single achievable aim. It's an exercise in the research process, after all.

Ronán Michael Conroy Cory Cobb

Thank you so very much for your response! My research question is: How do individuals of Middle Eastern origin conceptualize their sexual minority identity in relation to their cultural identification?

Per past studies, authors have raised the importance of minimizing the degree to which dominant frameworks of sexual minority identity development borrowed from Western culture are imposed upon Middle Eastern individuals.

Given this & the lack of empirical research on sexual minority individuals who are also Middle Eastern, we felt it would be appropriate to employ this kind of qualitative approach in order to explore the unique challenges they face as a whole.

However, I also wanted to follow up on that and look into the internalized homophobia & lack of disclosure of their sexual identity in comparison to that of the majority (white LGB individuals). I wanted to employ a covariance analysis to see if age, level of education, and resilience would affect these two constructs in Middle Eastern populations LGB vs White LGB populations.

Sorry if everything is all over the place as of now! As mentioned, it is my first independent undergraduate project, and I can appreciate any advice! Thanks so much in advance for your help :)

Hi Rawan, I must admit that I am still a bit unclear with respect to what you want to do, but I will take a shot at it. Assuming you are wanting to assess whether their are mean group differences between White LGB and non-Whites on internalized homophobia and lack of disclosure, while also assessing the role of age, education, and resilience, then you have a few options.

You could perform an Analysis of Covariance with groups as the IV and homophobia/lack of disclosure as outcomes while holding age, resilience, and education constant (controlling them).

You could also look at interactions in a regression framework such that you create interaction terms between groups and the covariates you referenced (age, education, resilience). However, with that may covariates of interest, you would have a four way interaction, which becomes difficult to interpret and requires sufficient power. I might just control for them instead.

Finally, you could regress the outcomes on the groups, save the difference scores as the outcome, and then predict those saved scores using age, resilience, and education. I would note a limitation here is the issue of reliability of difference scores.

There are many ways to do this, but the three methods above are common. I hope this helps.

Cory

Rawan Hedefa can I suggest that this is a research agenda rather than just a research project?

By this I mean that the insights you get from qualitative research are pretty much necessary to formulate the most appropriate measures for a quantitative phase. You will also find yourself with possible scales and measures that have not been validated on non-western populations. My sense would be that the second phase of the agenda would be doing a think-aloud study where the participants go through the potential quantitative measures. Questions that are socially appropriate and non-threatening in one culture can seem rude or intrusive in another. And you will also learn what things the questionnaires fail to tap by having people talk about the experience of answering them.

You have a really good area here. There is one more reason I would break it into stages, and that has to do with your professional development as a researcher. A really important question needs really top quality research. This means that you need to build up experience as a researcher and a feel for your area by using your first pieces of research very much as learning exercises. By taking on too much, you risk not learning and simply floundering with too much data that you can't put together.

I'd consider this research question the agenda for a PhD, frankly. And I'd really encourage you to pursue it like that, rather than trying to get it all done as an undergrad.

Does that make sense?