Dear all,
Thank you in advance for helping me choose the correct statistical test for my dissertation.
In my dissertation, I am investigating whether there is an association between marital status and type two diabetes.
The dependant variable is type two diabetes: ( yes, no)
The independant variable is marital status: (single, married, divorced, widowed)
The confounding variables are Age , Gended,ethnicity, BMI, Income, Education, socio-economic status, physical activity; all of them are categorical data.
I need help on which SPSS test shall I select.
Thank you
That´s depended on the categorical characteristis of the variables if they are quantitativo (t test) or qualitative (chi-square and so on)
The usual choice would be logistic regression. Marital status should be entered as 3 zero-one dummies. In a follow-up run, I'd consider interacting marital status with sex. Also possibly with age group. With all those categorical variables, you will need a lot of cases, since logit essentially populates a n-way table. It's unfortunate that you don't have BMI as a continuous variable. The lack of family history of diabetes will be a huge limitation of your analysis, means you are ignoring a known - and critical - explanator. See, for example, Article Diet, Lifestyle, and the Risk of Type 2 Diabetes Mellitus in Women
orArticle Incidence of Type 2 Diabetes in the Elderly German Populatio...
Yiu akso are missing information on smoking, which is a known risk factor. BMI also does not really account for diet or alcohol consumption.
Is not an association as you do not believe that diabetes produces marital status so we are in regression territory not correlation.
so as outcome is binary we are in logistic territory.
it now depends to some extent on how much data you have, if a lot I would match first then model. Search on matchit.
In addition to the other answers(Profs Miller and Jones) which are quite good, I assume this is a prediction study. Hence variable selection can give an idea of the relative importance of your predictors. For example, see the attached. Best wishes, D. Booth
- Badges
- Science topic
- Similar topics
- Medical Science Nursing
- Dissertations