I am undertaking a DEA analysis of banking in India but I am a bit confused about how to normalize data since some of my data is in Rupees some inputs are in number? Can anyone tell me that do I need to use GDP deflator for my analysis since I am comparing efficiency 2 year pre and post an event ?
@Kanvi A manuscript link below is about preparing the data for data envelopment analysis and might be really helpful to your cause. https://link.springer.com/chapter/10.1007%2F978-0-387-71607-7_17
Normalization will not change the results significantly because DEA is a ratio and normalization of numerator and denominator will not change results.
Normality of data is an important assumption for the purpose of applying parametric and non parametric test. However normality won't affect DEA as the data is in ratio scale. Since DEA is a ratio. Event study methodology is generally used to analyse the impact of pre and post event. Paired t test in parametric test is used for analysing the differences between two period. Whereas in non parametric test wilcoxon signed test an alternative for t test could be used.
Normalisation will not change your results using DEA analysis. The advice of Mayank will give you the most efficient results for your type of research.
Should not change your results because DEA ranks your data and the ranking is compared across different series of data you have.
Dear Kanvi.
Normalizing data will have no effect on DEA.
It's a data on a scale ratio already.
Regards
Check https://link.springer.com/article/10.1007/s10644-020-09281-w
Also check https://ideas.repec.org/a/ora/journl/v1y2019i1p107-113.html
Kindly check this Kandvi
https://www.khanacademy.org/economics-finance-domain/macroeconomics/macro-economic-indicators-and-the-business-cycle/macro-real-vs-nominal-gdp/a/adjusting-nominal-values-to-real-values-cnx
- Badges
- Science topic
- Similar topics
- Counseling Psychology
- Confusion