Hey folks,

I am running a difference-in-differences regression on a panel data set.

Consider the first box we get in the regression results. I attach a sample image here.

Array


According to Stata's tedidregress manual, on page 7, it states:

The first table gives information about the treated and control groups and about treatment timing.
The first segment with the title Group tells us the number of treated and control hospitals: 28 hospitals
were using the old procedure and 18 hospitals were using the new one. The second segment of the
table gives information about the first time we observe hospitals in the control group and the first
time we observe the treatment (the new admission procedure) for hospitals in the treatment group.
In this example, all hospitals that adopted the new procedure did so in April, time period 4. If some
hospitals had adopted the policy later, the minimum and maximum time of first treatment would
differ.
I believe I created my dummy variables for the treatment group and weeks correctly.

My questions are:

1. Does having 33 as Control Time Max mean that there are some hospitals in the Control group that begin to have data in week 33 instead of week 32?

2. If my interpretation in #1 is correct, should I use xtdidregree or didregress since I do not have a balanced and fixed panel data set.

Any thoughts?