I am doing frontier analysis of banks and I calculated the robust efficiency scores for each of the bank. Now I want to run a bootstrap truncated regression with efficiency scores as dependent variable to examine their relationship with characteristics of banks. How can I run the method proposed by simar and wilson (20007) in R. Please help in this regards.
Hi, maybe you can apply this model
https://www.researchgate.net/post/Can_anyone_guide_me_through_the_rDEA_package
Dominic Loske Thank you very much. I am already using this package however I am not able to add more than two environment variables because I have only 12 firm in the sample and the sample period is 10 years.I have calculated the bias corrected efficiency score through the " dea.robust " function i.e. includes no environment variable. Now I want to identify the determinants of the obtained efficiency score. Should I use OLS, GMM,TOBIT or the 2nd algorithm proposed by Simar and wilson(2007).
Yes I am using Simar and Wilson(2007), with 30 DMUs and 1000 repetitions.
For the example in https://analyticsdefined.com/data-envelopment-analysis-in-r/ with the fast food chains, I attach my model to this comment as a.txt file.
Let me know if this does not solve your problem :).
Dominic Loske Thank you very much, I have calculated the bias-corrected efficiency score using the " dea.robust " function which gives efficiency scores based on Simar and Wilson(1998) . Now I want to measure the efficiency scores using the Simar and Wilson(2007) algorithms. However I am not able to apply the "dea.env.robust" that gives efficiency scores based on Simar and Wilson(2007). The reason I have only 12 DMU's and the 5 environment variable in each year of sample period. Can you please provide me the R-code of Simar and Wilson(2007) algorithms especially the second algorithm i.e. bootstrapping truncated regression.
@Ashiq Mohd Ilyas have you got the R code? I am also looking for a bootstrap truncated regression code. Please let me know. Thank you.
A Latif Patwary You can use the "lt" function from this package https://cran.r-project.org/web/packages/truncSP/truncSP.pdf if you want to only use truncated bootstrapped regression.
However, if you want to execute simar and Wilson(2007,2011) two step bootstrapping algorithm with environment variables then use "dea.env.robust" from from https://cran.r-project.org/web/packages/rDEA/rDEA.pdf.
if you want to run a simple truncated regression then use "truncreg" function from https://cran.r-project.org/web/packages/truncreg/truncreg.pdf
Hope this will help you. All the best.
"Simar and Wilson(2007,2011) two step bootstrapping algorithm with environment variables". Thanks for this .. Bootstraps Truncated Regression.
Hi, anybody knows if I can add undesirables outputs from SBM-DEA in bootstrap truncated regression?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size