Hello everyone,
This is my first time using Statalist as well as first time trying to do anything with respect to RD design so please forgive me if I will not properly describe everything.
I am working on my thesis where I am trying to use RD method based on Average grades and optionally receiving honors at school (such as Cum Laude) and its effect on labour market outcomes.
I am trying to use an important check if the density of the running variable around the threshold. Lee & Lemieux (2010) and McCrary (2008) describe that regression discontinuity designs can be invalid if the running variable, in my case Average gradfe, can be precisely manipulated. A jump or fall in density at the threshold likely implies sorting which threatens the RD design. McCrary (2008) proposes a test based on an estimator for the discontinuity at the cutoff in the density function of the assignment variable. Cattaneo (2016) provides a manipulation tests following McCrary based on density discontinuity, constructed using the results for local polynomial distribution estimators - https://pdfs.semanticscholar.org/452...181701a659.pdf
I get the following Stata outputs (see below):
My question is. Based on the P value I receive from this test, also on the plot(graphical) outcome, is it plausible to reject the validity of RD design due to invalid (manipulated) running variable?
Thank you very much in advance for any useful thoughts on this issue.
This is my first time using Statalist as well as first time trying to do anything with respect to RD design so please forgive me if I will not properly describe everything.
I am working on my thesis where I am trying to use RD method based on Average grades and optionally receiving honors at school (such as Cum Laude) and its effect on labour market outcomes.
I am trying to use an important check if the density of the running variable around the threshold. Lee & Lemieux (2010) and McCrary (2008) describe that regression discontinuity designs can be invalid if the running variable, in my case Average gradfe, can be precisely manipulated. A jump or fall in density at the threshold likely implies sorting which threatens the RD design. McCrary (2008) proposes a test based on an estimator for the discontinuity at the cutoff in the density function of the assignment variable. Cattaneo (2016) provides a manipulation tests following McCrary based on density discontinuity, constructed using the results for local polynomial distribution estimators - https://pdfs.semanticscholar.org/452...181701a659.pdf
I get the following Stata outputs (see below):
My question is. Based on the P value I receive from this test, also on the plot(graphical) outcome, is it plausible to reject the validity of RD design due to invalid (manipulated) running variable?
Thank you very much in advance for any useful thoughts on this issue.
Comment