I have a group of controls (N=197) with a particular assay values. I am interested in testing my cases for the same assay and then comparing. I want to match at least 197 cases for age and gender with the controls before I start testing cases.
My initial response: Match first on gender, Then, within gender, match on age +/- 2 years. If you're feeling lucky, use age +/- 1 year. This may be sufficient, if the age distribution is similar between your cases and controls. If it yields too few matches, then consider opening the age range to, say, +/- 3 years, or +/- 4 years.
You'll probably get several instances where more than one case is a match for a given control. For these, you'll want to use a random assignment method to determine which case to use.
This sort of thing can be done in a number of computer packages for data management and statistics.
Thanks Mary ..I agree with your approach . Average age of my cases is less than 10 yrs average age of controls. The age distribution is not that similar.In that case I think I should consider age range of +/- 4 years first.
The Mayo Clinic has a free download for a SAS macro called gmatch. It allows for matching with multiple variables, gives options for size of range to match within, weights, and allows you to do 2:1, 3:1, etc matches for case/control. It takes only a little while to figure out how to use within your code but well worth the effort!
SAS 9.2 is not possible i think. But SAS 9 is possible. Try some surfing on net. But before install you should change your system time to 10 or 12 years back.
In case you or others are interested in age and gender matching case and control groups you might find the Matlab-programme below interesting. It randomly matches your groups. It let's you know how many of your subjects in the case group are matched/not matched yet and their 'priority' (i.e. how often can you match each individual if you match both groups randomly):