Omitted variable problem is a primary statistical challenge in various observational studies. Failure to control for the omitted variable bias in any regression analysis can alter the efficiency of results obtained. The purpose of this study is to compare the performance of four estimation methods (Proxy variable, Instrumental Variable, Fixed Effect, First Difference) in controlling the omitted variable problem when they are varying with time, constant over time and slightly varying with time. Results from the Monte Carlo study showed that, the prefect proxy variable estimator performed better than the other models under all three cases. The instrument Variable estimator performed better than the Fixed Effect and First Difference estimator except in the case when the omitted variable is constant over time. Also, the Fixed Effect performed better than First Difference estimator when the omitted variable is time-invariant and vice versa when the omitted is slightly varying with time.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:uu-412896 |
| Date | January 2020 |
| Creators | Amartey, Philomina |
| Publisher | Uppsala universitet, Statistiska institutionen |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.2126 seconds