Support #579
closed
Added by zoe young over 8 years ago.
Updated over 8 years ago.
Description
Hi I have merged 3 variables from wave3 to wave 5 they are : c_cgwri_dv, c_cgvfc_dv, and c_cgna_dv. I want to obtain the mean for these variables but only for those who are present in wave 5. I created 2 dummy variables where wave 5 =1 is present in wave 5 and wave 3=1 if present in wave 3. when i ran a cross tab it showed that the entire sample that were in wave 3 were in wave 5. Can someone confirm this as it seems strange that no attrition occured between wave 3 and 5?
Thanks Zoe
- Status changed from New to In Progress
- Assignee set to zoe young
- % Done changed from 0 to 50
- Private changed from Yes to No
Assuming you are Stata, some common minor mistakes that could have led to this problem:
(i) You coded the dummy variables=0 as default, so the dummy variables = 0 even for individuals not present in either wave instead of being "."
(ii) You did not do (i) but when looking at the cross-tab of wave3 wave5 dummies you did not use the missing option, that is, you typed
tab wave3 wave5
Instead of
tab wave3 wave5, missing
There is attrition between Waves 3 and 5 and there are new entrants in Wave 5 who were not present in Wave 3.
Best wishes,
Alita
- Status changed from In Progress to Closed
- % Done changed from 50 to 100
Also available in: Atom
PDF
Updated by 
Updated by Victoria Nolan