Bachelorand
New member
- Joined
- Apr 26, 2018
- Messages
- 1
Dear community,
for my thesis I analyse a have an unbalanced panel dataset of patients, where 200 out 5000 received the treatment in a given week in 2015.
However, the treatment starts in a random week and lasts for X consecutive weeks.
First I was thinking about propensity score matching here, but as I only have the age and gender for each patient, I figured that it won't be enough to this matching. Does anyone have an idea which matching mehtod would be possible here?
Second, is it possible to do a diff in diff estimation here, as my dataset is highly unbalanced? And what would I have to do.
The outcome is a count variable of interaction between caretaker and patient per week. So I want to see if while being treated, patients use the services lesser.
I use Stata by the way.
Any help would be much appreciated!
for my thesis I analyse a have an unbalanced panel dataset of patients, where 200 out 5000 received the treatment in a given week in 2015.
However, the treatment starts in a random week and lasts for X consecutive weeks.
First I was thinking about propensity score matching here, but as I only have the age and gender for each patient, I figured that it won't be enough to this matching. Does anyone have an idea which matching mehtod would be possible here?
Second, is it possible to do a diff in diff estimation here, as my dataset is highly unbalanced? And what would I have to do.
The outcome is a count variable of interaction between caretaker and patient per week. So I want to see if while being treated, patients use the services lesser.
I use Stata by the way.
Any help would be much appreciated!