My research is about patient safety culture among healthcare workers using the Hospital survey on patient safety culture (HSOPSC). I am performing a multiple linear regression to predict the outcome "overall perception of patient safety" score composite from the independent variables which include the other patient safety culture composite scores (11 variables) and sociodemographic factors (9 variables). I can say that the sociodemographic factors can be potential confounders but I need to make sure. How do I perform that using SPSS? and after that, how to run MLR adjusting for confounders? Thank you
It sounds like the outcome variable is a composite that is made up (though perhaps only in part) of the 11 other safety variables you are using to predict that outcome. Am I understanding that correctly?
If so, the 11 predictors should yield an R-squared of 1.0 (or at least very close to 1.0, if the outcome composite involves more than just those 11). That would mean you need to re-think your approach to the analysis, as it does not make much sense to predict a composite score from the elements making up the composite. There will be no (meaningful) variance in the outcome left for the demographic variables to cover, once the first 11 variables are in the model.
What is your central research question? What relationship(s) are you looking for that you think might be confounded by demographic characteristics?
Sorry for not answering your question directly, but my advice would be to draw before analyzing. In order to know which variables to include in my model, based on the confounder criteria, I would first draw a directed acyclic graph based on evidence from the litterature to find out which variables to include in my model, and which variables are e.g. mediating factors, which should not be included in the model:. Check for instance:
Causal diagrams for epidemiologic research by Greenland et al.
and
https://cran.r-project.org/web/packages/ggdag/vignettes/intro-to-dags.html
A multiple linear regression model will not tell whether are variable is a mediator, and whether collider bias has been introduced by adjusting for a certain variable.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude