search

xuyiqing / gsynth

Generalized Synthetic Control Method
Other
131 stars 40 forks source link

No feature to estimate overall average effect over whole period for subgroups using cumuEff? #60

Open pbritteon opened 3 years ago

pbritteon commented 3 years ago

I have noticed that there is no feature to estimate the overall/average effect over whole post-policy period for subgroups using cumuEff package.

I was wondering whether this is something that might be added in the future or whether there is a way of calculating this from the est.catt output?

I think this would be a really nice feature as as a single estimate of the average effect per treatment unit is easier to present than trying to summarise results over time for each treatment unit in journals (when limited by space/figures).

P.S. Thank you for putting together such a great package

xuyiqing commented 3 years ago

Thank you for your kind words. We'll take notes. We're in the process of creating a new package, "fect", of which "gsynth" is a special case. We'll consider incorporating that.

To answer your question, yes, you can simply calculate a cumulative effect using each bootstrap draw and use the percentile method to get the confidence intervals.

Best. Yiqing

On Wed, May 26, 2021 at 1:48 AM pbritteon @.***> wrote:

I have noticed that there is no feature to estimate the overall/average effect over whole post-policy period for subgroups using cumuEff package.

I was wondering whether this is something that might be added in the future or whether there is a way of calculating this from the est.catt output?

I think this would be a really nice feature as as a single estimate of the average effect per treatment unit is easier to present than trying to summarise results over time for each treatment unit in journals (when limited by space/figures).

P.S. Thank you for putting together such a great package

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/xuyiqing/gsynth/issues/60, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB2PKGHYNA7R6MGXRBM52PTTPSYVRANCNFSM45RQALRA .

pbritteon commented 3 years ago

Thanks for your quick response, Yiqing. I will look into fect / keep an eye out for developments.

Ok thanks! Is there a package in R that will allow me to do this easily? I am having to get use to R having used stata for years so any pointers would be much appreciated

xuyiqing commented 3 years ago

No problem.

Ziyi, could you help? I know you're very busy recently. Maybe we shall add it to the new version.

On Wed, May 26, 2021, 02:26 pbritteon @.***> wrote:

Thanks for your quick response, Yiqing. I will look into fect / keep an eye out for developments.

Ok thanks! Is there a package in R that will allow me to do this easily? I am having to get use to R having used stata for years so any pointers would be much appreciated

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/xuyiqing/gsynth/issues/60#issuecomment-848616890, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB2PKGGEUHOC2WC6X5KFHWDTPS5FTANCNFSM45RQALRA .

Katebellis commented 1 year ago

Hi all,

I was wondering whether there has been any progression related to this query? Or whether you could elaborate more on how to do this outside of the package functions please Yiqing?

I've had a look at the fect package and dont believe I can reproduce gsynth results using IFEct as they're using different algorithms? But please correct me if I'm wrong.

Many thanks in advance