inference randomization

[Yannis Cer]

Hi everyone, I hope you are all doing good. I have to do an inference randomization but I don't really understand and know how to proceed, could someone help me ? For a variable I have to test the hypothesis that there was no treatment effect using randomization inference. I have to use 400 random reassignments to get the empirical distribution of the treatment effect estimator. I don't understand first how to write the code and second I don't get how can I see the treatment effect on the new treated groups ? Like a data base only provides treatment effect observation for the first group assigned to the treatment no ? I hope you understood what I don't understand and some of you help me out with it. Thank you all in advance

[Carlo Lazzaro]

Yannis:
welcome to this forum.
I find the description of your query unclear.
That said, the Stata commands that spring to my mind are: -permute- and -eteffects-.
As an aside, if your query relates to an assignment, please consider https://www.statalist.org/forums/help#adviceextras #4. Thanks.

[Yannis Cer]

First of all, thank you for your answer. My purpose here is to understand how this concept works, and then I will figure it out by my own on how to do it I don't want answers. I just don't really get this process of inference randomization haha, but anyways I will keep searching. Have a lovely day,

[Yannis Cer]

For me, randomization inference: "considers what would have occurred under not only the random assignment that happened to be selected for the experiment, but rather under all possible random assignments: would the results hold?" (Source: Randomization Inference - DIME Wiki (worldbank.org)) this part I understand. I just don't get how do we can have the results of the new treated individuals because we don't observe it ?

[Carlo Lazzaro]

Yannis:
take a look at -permute- entry, Stata .pdf manual.