Consider the below table of data, which is from a context in which one group got...

Transcribed Image Text:

Consider the below table of data, which is from a context in which one group got a treatment between time 2 and time 3, and a different group didn't. This table contains two versions of the Y variable: Y, which is the actual outcome we can observe, and Y*, which is the outcome we would have observed if the treatment had never occurred. Group Treated Treated Treated Control Control Control Time 1 (Before) 2 (Before) 3 (After) 1 (Before) 2 (Before) 3 (After) Y 1 12 5 3 4 Y* 1 12 3 3 4 6 1. Does parallel trends hold here? Show the calculation that you perform to check whether parallel trends holds. (Note: we're checking parallel trends here, not prior trends. We can check parallel trends directly since we have Y*) 2. Using the real observed data Y, calculate the difference-in-differences estimate of the effect of treatment, using period 3 as After and period 2 as Before. Consider the below table of data, which is from a context in which one group got a treatment between time 2 and time 3, and a different group didn't. This table contains two versions of the Y variable: Y, which is the actual outcome we can observe, and Y*, which is the outcome we would have observed if the treatment had never occurred. Group Treated Treated Treated Control Control Control Time 1 (Before) 2 (Before) 3 (After) 1 (Before) 2 (Before) 3 (After) Y 1 12 5 3 4 Y* 1 12 3 3 4 6 1. Does parallel trends hold here? Show the calculation that you perform to check whether parallel trends holds. (Note: we're checking parallel trends here, not prior trends. We can check parallel trends directly since we have Y*) 2. Using the real observed data Y, calculate the difference-in-differences estimate of the effect of treatment, using period 3 as After and period 2 as Before.

Answer rating: 100% (QA)

To check whether parallel trends hold in the given data we need to compare the trends of the treated ... View the full answer