need help drawing matched controls group

[hopelessatstats]

Hello, I am doing a study on schizophrenia and cannabis use. I currently have a group of 35 schizophrenia patients who have a history of cannabis use and need to recruit a matched healthy normal control group for comparison.

I was wondering if u had any information on drawing a matched control group using z-scores? And how one can calculate this easier on spss rather than by hand? Although I'm willing to do this by hand if needed. I must create a matched control group on the variables: age, gender, IQ, years of education and cannabis use. I'm terrible at stats!

 

[Link]

ok, your study population is persons with schizophrenia. Your exposure is cannabis use. What's your outcome? What are you trying to determine?

I'm a little confused by what you would do with the z-scores. In matched studies, you would compare rates in the exposed group to rates in the matched unexposed groups.

 

[hopelessatstats]

I'm looking at the combined effects of schizophrenia + cannabis use on cognitive function. It would have been much easier to compare the performance of this group with the performance of individuals with schizophrenia who do not have a drug use history. However, we have been unsuccessful in being able to find an adequate amount of these people. Therefore, we wanted to compare the performance of the schiz + cannabis use group with a group of healthy normals who have a similar history of cannabis use and are matched for age, gender, IQ and years of education. I was thinking I needed to use z-scores to make sure my recruited matched controls were appropriately matched on all of these variables.

 

[jeromyanglim]

Check out this article
* Rosenbaum, P.R. and Rubin, D.B. 1985: Constructing a control group using multivariate matched sampling methods that incorporate the propensity score
* PDF seems to be available here

The article provides an in depth discussion of issues involved with creating matched samples.

In addition, think deeply about the implications of treating IQ as a control variable. It is quite likely that if cannabis negatively affects cognitive function, then it would also negatively affect IQ.
In fact, one loose definition of IQ is cognitive functioning.
You don't want to control for variables that are actually part of the effect of interest.