Hi, So I have longitudinal data from two time points. Data set 1 has 39 variables with 1707 observations and data set 2 has 24 variables with 1707 observations. I am supposed to merge these data sets so that the observations can be traced back to one individual and then analysed. The data is de identified which i think it causing the issue in merging. IDs have been assigned to each individual. I merged the data using an m:1 merge via the IDs and got this result :
Result Number of obs
-----------------------------------------
Not matched 2,844
from master 1,422 (_merge==1)
from using 1,422 (_merge==2)

Matched 285 (_merge==3)
It appears that the merge was successful but the observations did not match? Is there any reason this could be happening? does this mean the observations are not linked or is this merged data set valid. The new merged data set has given a bunch of missing values for the unmatched obs. Do i adjust for this and use the merge data set?

Another idea was to match all the variables in both sets that are common since it is de identified data so the ID labels are not the same. But when I do so it doesnt work for some reason.
Any help or suggestions are highlt appreciated!