Hi all, we have a data set with 258 records (84 cases , 174 controls). We did frequency matching for age and sex during the data collection , however data entry did not involve a matched (link) variable to link each case to two specific controls. Now we want to carry out a matched analysis and need to define a new link variable, is there a way to generate one from the existing data about case/control status, 10 year age categories and sex.I already defined a new "sex age" variable so there are about 16 different strata (8 age categories and sex) but I am not sure if it is of any help. Can someone please help ?
Your situation is complex. Why you did not enter the matching variable?
Can you retrieve a which ID case is linked to which ID control? If Yes. Enter that in a excel file and merge this excel dataset.
For example you could make the file look like the following columns:
ID IDCASE
1 1
19 1
67 1
7 7
74 7
69 7
...
The first collumn is for each ID on your current dataset. And the second collumn (IDCASE) is what ID is the case. Thus, (i) no duplications on ID collun and (ii) when ID and IDCASE are equal it is case otherwise is a control. In this example 1 and 7 are cases. The rest are controls.
Then you may merge this excel file on EpiInfo or other software.
Hope this helps
- Badges
- Science topic
- Similar topics
- Filing