Parallel trend assumption

[ellenjunghyunkim]

Thank you for your package!

All my questions are related to parallel trend assumption....!

How can we change the baseline outcome comparison from Y_(t-1) to Y_(t-2)?
This is to address 1-period anticipation before the treatment.
And as someone already raised here, it would be wonderful to have the outcome values themselves in addition to seeing the standard difference.

Kind regards

[insongkim]

@ellenjunghyunkim Thanks for using our package.

We plan to include an option for users to conduct placebo tests, e.g., comparing Y_{t-1} vs Y_{t-2} as you suggested. We will let you know when such functionality becomes available.

[ellenjunghyunkim]

Thank you for your comment!

My question was toward the direction of allowing some anticipatory effect.
When the treatment is a choice made by an individual, allowing some anticipatory periods makes sense to me.

To allow k-period anticipation, instead of having Y_{t+F} - Y_{t-1}, can we change the base pre-treatment year to Y_{t-(k+1)} to have Y_{t+F} - Y_{t-(k+1)}? This is to compare the future outcomes to the year when we are free from anticipatory effects. It would be nice to have this functionality!

Kind regards,