We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
An Exercise in Cross-Validation of Data on Employment and Unemployment.
- Authors
NIRAD, Inamdar
- Abstract
This paper compares two well-known resources of socioeconomic data of India -- one from National Sample Survey Office (NSSO) 68th round and one from India Human Development Survey 2nd round (IHDS-II). Academic research hinges on validation of underlying data, especially in the socioeconomic field, since it can have Government policy implications. This study compares the two data sources from the same time frame (2011 - 2012) on relevant parameters pertaining to employment and unemployment. In particular, for the greater part of the study, we focus on women. First, we identify some factors influencing the labour force participation rate (LFPR) and stock of persons' unemployment rate (SPUR) of women -- sector, marital status, education, relation to household head, religion, social group and household size. We perform a series of Probit analyses these separately to determine which factors are significant for LFPR and SPUR. Then, we combine and include their interaction effects. At every step, we compare the results from NSSO and IHDS-II to cross-validate them. We find that for LFPR, both data sources give similar results that all variables are significant, but for SPUR, they give different results. Finally, we perform Kolmogorov- Smirnov (K-S) tests and observe that both samples are normally distributed. However, given the disparity of results in a majority of analyses, one cannot say that the two data sources are similar. Therefore, in the absence of cross-validation of data from NSSO & IHDSII, the scope for further research is to include a third data source and derive results from similar analyses. Finally, a comprehensive study would entail proposing a consolidated model, based on three sources.
- Publication
Annals of the University Dunarea de Jos of Galati: Fascicle: I, Economics & Applied Informatics, 2019, Vol 25, Issue 3, p72
- ISSN
1584-0409
- Publication type
Academic Journal
- DOI
10.35219/eai1584040957