What Type Of Research Design is This?

[marblegatherer]

Oy. My research design methodology-fu is severely lacking today.

If I were designing a study where I looked at participants' recalled fantasies of adolescence and the correlation between those fantasies and current adult functioning on several measures, what would this design type be called?

Thanks for any help, the more I read, the more I get confused.
*sigh*
Why can't everything be either theory or direct application?

---

Members don't see this ad.

 

[cognosco]

If your variables are continuous, and you are simply trying to show that there is a relationship between number/type/content/whatever of fantasies during adolescence and subsequent behavior during adulthood, you are running a correlational study.

If you are trying to demonstrate that a relationship exists and also that fantsy predicts subsequent behavior, you are running a regression. If you are using multiple aspects of the IV (fantansy), you are running a multiple regression.

If you are trying to say that fantasy causes subsequent behavior, you are running an ANOVA. If you have covariates, it's an ANCOVA. If you have multiple DVs (i.e. if you are saying that fantasy causes changes to several different variables), you are running a MANOVA. Hope this helps.

 

[marblegatherer]

Thanks, cognosco! That helps a lot.

I am trying to say that certain attributes of fantasies (amount of ego diffusion present) predicts later adjustment (fullfillment or distress levels), their adult level of ego development, and social role.

So I guess it is a multiple regression study? Nonexperimental design?

 

[Ollie123]

"Multiple regression study" is not a research design, its an analytic technique.

It is definitely not experimental. You aren't manipulating anything.

I'd stick with either correlational or survey to describe the study. It might also be important to mention that its retrospective, since that is obviously a gigantic limitation.

 

[JockNerd]

Oy. My research design methodology-fu is severely lacking today.

If I were designing a study where I looked at participants' recalled fantasies of adolescence and the correlation between those fantasies and current adult functioning on several measures, what would this design type be called?

Thanks for any help, the more I read, the more I get confused.
*sigh*
Why can't everything be either theory or direct application?

Thanks, cognosco! That helps a lot.

I am trying to say that certain attributes of fantasies (amount of ego diffusion present) predicts later adjustment (fullfillment or distress levels), their adult level of ego development, and social role.

So I guess it is a multiple regression study? Nonexperimental design?

Your research design and your analysis method are two different things. You're doing a correlational, cross-sectional design. How you chose to analyse it doesn't let you infer causation about anything.

Also, from your described design, you're not really looking at the attributes of childhood fantasies, are you? You're looking at recalled fantasies, which is different. It's pretty reasonable to guess that people with better ego development and a stronger sense of their social role would recall happier/better adjusted/more relational/whatever fantasies.

I don't mean to be overly critical, but if you tried to publish that paper framed as a predictive study that assessed childhood fantasy, rather than a correlational study that looked at recalled fantasy, it wouldn't fly anywhere.

 

[KillerDiller]

Did you assess ego development with the WUSCT? Just curious, as that is what I used for my thesis.

 

[marblegatherer]

Did you assess ego development with the WUSCT? Just curious, as that is what I used for my thesis.

Haven't yet, but planning on using EOM-EIS.

What was your thesis topic? And might I have a peek at your reference list?

 

[KillerDiller]

Haven't yet, but planning on using EOM-EIS.

What was your thesis topic? And might I have a peek at your reference list?

My thesis looked into developmental characteristics and adolescent sexual behavior. I don't think my reference list would be much use to you, though, as I just had to google EOM-EIS to see what it was. I'm pretty sure we're working under completely different models of ego development.

 

[cmuhooligan]

If you are trying to say that fantasy causes subsequent behavior, you are running an ANOVA. If you have covariates, it's an ANCOVA. If you have multiple DVs (i.e. if you are saying that fantasy causes changes to several different variables), you are running a MANOVA. Hope this helps.

ANOVA, ANCOVA, MANOVA do not necessarily infer causality--it depends on the type of design you're conducting.