Convert odds ratios and beta coefficients to a comparable effect size

#1
Dear all,

First a short introduction to my study/review/meta-analysis.

The purpose of my study is to systematically review longitudinal evidence on the association between psychosocial work characteristics (i.e. job demands, job control, social support) and sleep, since cross-sectional research indicates a link. My main research question is if psychosocial work characteristics can predict sleep quality.

An extensive literature search resulted in 10 longitudinal studies, in which the work characteristics are the independent variables and (poor) sleep quality is the dependent variable. All variables were measured by questionnaires and most of the time comparable instruments were used. 8 of the studies performed logistic regression analyses and outcomes were odds ratios for job demands, job control and social support separately. The 2 other studies performed Structural Equation Modeling and only reported beta coefficients.

My question is if and how I will be able to convert these effect sizes to one comparable effect size.

I'm looking forward to your suggestions and hope you can help!

Already many thanks in advance!

Kind regards,

Michelle
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Ideally, if the longitudinal studies were prospective they should report relative risks not odds ratios? Without knowing the studies, you can always convert beta coefficients into odds ratios if you also have the standard errors. But I am not well versed in SEM, so I don't know if it would be an appropriate comparisons.

I am also guessing these are adjusted odds ratios, controlling for the other variables, correct?