new diffindiff slides

[gustavekenedi]

Hi Florian,

Here are our proposed slides for the new diff in diff lecture. I apologize in advance for having committed other files so you'll see 24 files changed, I hope this won't be too much of a hassle :/ Other than the .Rmd, there's only one other relevant image file: nj_penn_map.png.

For this lecture, I've taken the example of Card and Krueger (1994) on minimum wage's impact on employment to illustrate the DiD method. I've included two small tutorials for students using the actual data from the study to replicate some of the basic findings. I've tried to illustrate as best as possible how DiD works graphically and what the common trends assumption means.
Let me know what you think.

At the end of the lecture I have two slides for more advanced topics: multiperiod multistate extensions of DiD and standard error clustering (just the intuition, not the maths). I don't think it's necessary though the clustering may get students thinking about standard errors and not just point estimates which is important.
I assume we'll have to create a quiz with some questions on DiD. Duflo's (2001) paper (https://economics.mit.edu/files/726) could be used (other suggestions can be found here: https://twitter.com/i/moments/994859975968280576?lang=en).

Thanks!

I'll upload the new RDD slides very soon.

Gustave

[floswald]

great! I will be able to look tomorrow or wednesday. just as a further idea, it might be cool to use those animations of nick nk. i just made a fork here https://github.com/ScPoEcon/causalgraphs checkout the one on DiD which is a video here

all of his animations are cool, so let's try to use as many as we can!

[gustavekenedi]

Thanks for the suggestion! We could make those animations with the real Card and Krueger (1994) data :)

[floswald]

Good idea!

…

On Mon 20 Jan 2020 at 22:21, Gustave Kenedi ***@***.***> wrote: Thanks for the suggestion! We could make those animations with the real Card and Krueger (1994) data :) — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#6?email_source=notifications&email_token=AAHZM5RNT4KWLUNLPCSIIIDQ6YIWRA5CNFSM4KJH6RVKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJN25DI#issuecomment-576433805>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAHZM5XUUMEXOZ2UMOD3M3LQ6YIWRANCNFSM4KJH6RVA> .

[gustavekenedi]

A new update to remark.js has messed up how some of the slides are rendered (in particular, there is now a horizontal and/or vertical scrolling bar). The solution (as discussed here) appears to revert to the previous version of remark.js by adding the following code:

output:
  xaringan::moon_reader:
    chakra: "https://cdnjs.cloudflare.com/ajax/libs/remark/0.14.0/remark.min.js"

It's probably not the ideal solution but it works.
Let me know if you also encounter the same issue and if this fix works on your end as well :)

[floswald]

ok. linking to the previous version of remark.js seems the best solution to me at the moment (I have not verified, but it seems clear there is an issue)

…

On Tue, 21 Jan 2020 at 12:23, Gustave Kenedi ***@***.***> wrote: A new update to remark.js has messed up how some of the slides are rendered (in particular, there is now a horizontal and/or vertical scrolling bar). The solution (as discussed here <yihui/xaringan#245>) appears to revert to the previous version of remark.js by adding the following code: output: xaringan::moon_reader: chakra: "https://cdnjs.cloudflare.com/ajax/libs/remark/0.14.0/remark.min.js" It's probably not the ideal solution but it works. Let me know if you also encounter the same issue and if this fix works on your end as well :) — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#6?email_source=notifications&email_token=AAHZM5XW7RZMN2N5WXQQ55LQ63LLFA5CNFSM4KJH6RVKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJPM6MA#issuecomment-576638768>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAHZM5VEMSTVOH43WSKXSATQ63LLFANCNFSM4KJH6RVA> .

[floswald]

Some more inspiration maybe
https://blogs.worldbank.org/impactevaluations/revisiting-difference-differences-parallel-trends-assumption-part-i-pre-trend?cid=SHR_BlogSiteShare_EN_EXT

[floswald]

beautiful, absolutely love it. those graphs are stunning. well done!

just some comments to work on. and please clean up the included files (there is some stuff from chapter 1 committed), so it will be difficult to merge. i left a comment how to do it.

[gustavekenedi]

Thanks for all the feedback and suggestions! I'll go over them this weekend (and take care of the git issues)!

[floswald]

bumping that PR to see what's going on

[floswald]

few minor comments. overall extremely cool set of slides. Still love the graphs. good work.

[gustavekenedi]

Thanks for all your feedback and suggestions! @pvilledieu may make some other final changes to some of the graphs so let's keep the PR open just in case if you don't mind :)

[floswald]

so this is good to merge?

[pvilledieu]

Yes it’s good !

[gustavekenedi]

so this is good to merge?

yes! thanks!