Which type of analysis should I do? HELP!! Dissertation Issues!

[donthatekate]

Finally getting back to research for my dissertation after a 4 year break from the classroom!! need some help on which analyses are best for some of the following scenarios.

Thanks!!

1-test the statistical significance of and type of relationship between a DV (scale level of measure—i.e. 0=no use at all, 4=very likely to use) and IV (scale level of measure—i.e. 0=no knowledge, 4=advanced knowledge)

2-determine if any of the control variables (i.e age/race/gender/years of experience) impact the relationship between the IV and the DV (considering control variables are at the the nominal and ordinal levels of measure)

3-determine if any of the control variables (nominal and ordinal levels of measure) directly impact or show any patterns with the DV (scale level of measure) or IVs (scale levels of measure)

 

[CowboyBear]

Hi there, welcome! I would suggest:

1 - spearman's rho could be useful here

2 - I wouldn't do this - atheoretically hunting for interactions between the effects of your independent variables and demographic controls is not really a useful thing to do, and risks high rates of Type 1 error.

3 - include control variables in your main models if there's a substantive reason to expect they're potential confounding variables. Don't include/exclude based on their relationships with IVs or DV (i.e., if you don't really need to do 3).

PS. Your "scale" measures look ordinal to me?