Abstract: The purpose of this study is to analyze the relationship between skill improvement, and the environments of companies in Korea using HRD-Index, which quantifies levels of HRD investment, activities, and the environments of companies in Korea. The OLS regression analysis cannot eliminate the selection bias, and to address this issue, outcomes were compared between an OLS regression analysis and an analysis with the GPS-based IPW. Covariates affecting HRD-Index were selected, GPS was estimated and the relationship between HRD-Index and skill improvement was analyzed using IPW. Based on data obtained from the 2017 Human Capital Corporate Panel (HCCP) in Korea, 337 responses were analyzed. Selected covariates were the number of workers, high-educated ratio, the core talent definition, the HR plan, the HR plan-management strategy linkage, job analysis, and sales, and these were used to determine point estimates and 95% confidence intervals. Companies with the bottom 10% of HRD-Index scores were found to have lower skill improvements than in the OLS regression analysis, but companies with the bottom 40% of HRD-Index scores, i.e., 44 points or higher, were measured that the point estimation value applying the IPW is larger than the point estimation value of the OLS regression analysis. It means that the ordinary OLS regression analysis led to underestimations. Hence, it can be said that existing studies in Korea have underestimated the effect of HRD, and the higher the level of HRD, the greater the effect than existing perceptions.
