Talk:Regression discontinuity design

Discrete Data

It would be nice to include a short section on RDD when the assignment variable is discrete. This would essentially bring up the points discussed in Lee and Card (2007). I'll try to get around to this, but as you can tell, i'm coming up with ideas faster than I can write them up. --PatrickButton (talk) 04:25, 14 April 2012 (UTC)

Examples

It would be a great idea to add more examples of RDD from a range of disciplines. A really good list is in Lee and Lemieux (2010). I'll try to put some of those in here later. --PatrickButton (talk) 09:46, 21 March 2012 (UTC)

reference completion

The article has two references whose details need completion. Melcombe (talk) 16:55, 30 April 2008 (UTC)

Campbell, Cook and Shadish discuss RD designs as quasi-experiments

Campbell, Cook and Shadish discuss RD designs as quasi-experiments (among the best observational study designs). Calling these studies "experiments" (sic.) is unorthodox and perhaps heretical! Could we have some page references, please? Kiefer.Wolfowitz (talk) 22:25, 6 June 2010 (UTC)

RDD is commonly called a quasi-experiment in economics. Instrumental variables and difference-in-differences (or similar panel data methods) are also considered quasi-experiments. I agree that the term is controversial but it's very common. I'll see what I can dig up in terms of references for that. --PatrickButton (talk) 09:46, 21 March 2012 (UTC)

Further additions needed

- Mathematical argument => Add some maths to illustrate the intuition in a formalized way. -I'll try to add this once I figure out how to do equations in Wikipedia. --PatrickButton (talk) 09:46, 21 March 2012 (UTC)

- Improve Advantages/Disadvantages => This is a horribly flawed section - should improve.

- Extensions => Fuzzy discontinuity design

--Guo (talk) 11:32, 16 July 2011 (UTC)

Dr. Fujii's comment on this article

Dr. Fujii has reviewed this Wikipedia page, and provided us with the following comments to improve its quality:

It may be useful to add the point that the before-after analysis can be seen as a special case of RDD (where the discontinuity occurs along the time dimension). This helps the readers to understand how RDD works and why the contamination issue discussed in "Disadvantages" is often problematic in before-after analysis. See Lee, MJ (2005) "Micro-econometrics for policy, program and treatment effects" Oxford University Press.

We hope Wikipedians on this talk page can take advantage of these comments and improve the quality of the article accordingly.

Dr. Fujii has published scholarly research which seems to be relevant to this Wikipedia article:

• Reference : Tomoki Fujii, 2011. "Impact of food inflation on poverty in the Philippines," Working Papers 14-2011, Singapore Management University, School of Economics.

ExpertIdeasBot (talk) 16:53, 19 May 2016 (UTC)

Dr. Stancanelli's comment on this article

Dr. Stancanelli has reviewed this Wikipedia page, and provided us with the following comments to improve its quality:

1. e.g. scholarship award)

I would here add references to the study of RD with scholarship award:

van der Klaauw, Wilbert (2002), Estimating the Effect of Financial Aid Offers on College Enrollment: A Regression-Discontinuity Approach, International Economic Review, 43(4), 1249-1287.

2. h {\displaystyle h} h is the bandwidth of data used.

Here I would also say that h is the distance from the cut-off, so I would write:

h {\displaystyle h} h is the bandwidth (or chosen distance from the cut-off, on the two sides of the cut-off) of data used.

2. Parametric estimation.

I am not so sure what happened to parametric estimation as this seems still empty. I would suggest the following:

Parametric estimation by means of linear regressions that include a dummy for observations above the cut-off and (interactions with) a polynomial in the running variable among the explanatory variables is essentially equivalent to non-parametric estimation, at least for a given bandwidth and appropriate choice of the degree of the polynomial in the running variable.

References

3. In contrast to the sharp regression discontinuity design, a fuzzy regression discontinuity design (FRDD) does not require a sharp discontinuity in the probability of assignment but is applicable as long as the probability of assignment is different.

I think this can be improved as:

In contrast to the sharp regression discontinuity design, a fuzzy regression discontinuity design (FRDD) does not require a sharp discontinuity in the probability of assignment but is applicable as long as the probability of assignment is different from zero, and in particular greater than zero (though less than one).

Hahn, Jinyong; Petra Todd; Wilbert Van der Klaauw (2001), Regression-Discontinuity Design, Econometrica, 69 (1), pp. 201-209.

We hope Wikipedians on this talk page can take advantage of these comments and improve the quality of the article accordingly.

We believe Dr. Stancanelli has expertise on the topic of this article, since he has published relevant scholarly research:

• Reference : Elena Stancanelli, 2012. "Spouses' Retirement and Hours of Work Outcomes: Evidence from Twofold Regression Discontinuity," Documents de travail du Centre d'Economie de la Sorbonne 12074, Universite Pantheon-Sorbonne (Paris 1), Centre d'Economie de la Sorbonne.

ExpertIdeasBot (talk) 16:38, 2 August 2016 (UTC)