I have seen a couple of panel data research where the authors fill up missing data of a variable with similar variable from another source. For example , getting inequality data from WDI and filling the missing values up with data from SWIID. Is there any justification for this and also how do we fill this up when the 2 estimates are Calculated differently and with different scales.
Dear Hamid Muili you can use interpolation method as suggested by Ette Etuk .
Hamid Muili A more appropriate way to "fill in" missing data (more appropriate than any single-imputation method) is by multiple imputation. With multiple imputation, you can borrow information from multiple sources ("auxiliary variables") in the imputation process and you do this more than once (obtaining multiple imputed data sets across which statistical results are averaged or "pooled"). For details, see
Enders, C. K. (2010). Applied missing data analysis. New York: Guilford Press.
Hamid Muili you should not fill up missing data from different sources. You may either go with interpolation as suggested by Mohammad Razib Hossain and Ette Etuk or you may consider the regression imputation method as suggested by Christian Geiser for more accuracy. You may also go through the following paper,
Kang H. (2013). The prevention and handling of the missing data. Korean journal of anesthesiology, 64(5), 402–406. https://doi.org/10.4097/kjae.2013.64.5.402
Dear Hamid Muili
You can also look for substitutes for your variables if you want to avoid interpolation
Dear Humid.
The use of interpolation is better than filling missing data from one source with another source. The units and other conditions surrounding the data collection is different.
However, you can make use of proxy, that is using other close related variables to proxy the missing variables as a whole.
Regards
Actually analysis is applied with the planned variables beforehand. So dataset reflects the relations amongst variables in the dataset. The term term ceteris paribus, error terms and assumptions are closely related to the purpose of the study. A different variable means a new data set and leads to the differents structural relations. Because the predetermined analysis is not for this new data structure, using proxy or some other variables makes a negative effect on estimation process. That's why interpolation or some deletion (listwise or something else) can be used.
- Badges
- Science topic
- Similar topics
- Chemistry
- Biochemistry