I am doing a censored bivariate probit with sample selection
So, I am trying to estimate the probability for an individual to be hired in the private sector. In doing so, I have to include the probability of participating in the labour market to account for sample selection by having 2 stage analysis.
The equations are as the following:
Males sample:
1st stage, Selection equation, Labour force participation (1 if in the LF, Otherwise, 0)= individuals charchetristics such as ag education, region, nationality + Instrument
2nd stage, outcome equation, Probability to be hired in the private sector(1 if hired in the private sector, Otherwise, 0)= individuals charchetristics such as ag education, region, nationality
Females sample:
1st stage, Selection equation, Labour force participation (1 if in the LF, Otherwise, 0)= individuals charchetristics such as ag education, region, nationality + Instrument
2nd stage, outcome equation, Probability to be hired in the private sector(1 if hired in the private sector, Otherwise, 0)= individuals charchetristics such as ag education, region, nationality
The question now is, how to perform the t-test between these 2 sample (Males-Females) after running the regression to compare the changes in the coefficients.
Thank you
t-test is designed to compare regression factors. Make sure that your regression factors are applied or not on normal results or not: test before or the normality of your results and then apply a Pearson or a Spearman on each f or m serie. Then t-test will tell you how much your results are likely to rely on random or to be reproductible on other samples.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running