Do imputed education histories provide satisfactory results in fertility analysis in the Western German context?



Zabel: Imputed education histories and fertility analysis in the western German context

not have a child, time actually spent holding a lower level degree is more likely to be
lost in the imputed histories (Figure 1). In effect then, exposure time belonging to
people who did not have a child is removed from the ‘lower degree’ category in the
imputed histories. This of course raises the estimated risk of first birth for the ‘lower
degree’ category when using the imputed, as compared to the original, histories.

The problem thus is that imputations of the education variable, which is our
predictor variable, depend on educational attainment at a later point in time, which in
turn may have already been influenced by our dependent variable, the transition to the
first child. As described above, we would expect the consequence to be an
overestimation of transition rates to the first child for the status ‘lower level degree, not
enrolled in education,’ because people who did not have a child are selectively taken
out of this status. This problem is similar to the problem discussed formally by Hoem
and Funck Jensen (1982). The authors analyze how bias is caused in fertility analysis
when respondents are considered to be holding their highest educational degree
throughout the entire risk period. In this study, we will empirically investigate the
consequences of imputation when respondents are not considered to be holding their
highest degree the whole time, but only from the date they obtained their highest degree
onwards. Previous to this date, however, we consider them to be enrolled continuously.
We will investigate to what extent this procedure still entails bias.

Figure 1:   Reenrollment at a later point in time: imputed and original

education histories

high degree, not enrolled
low degree, not enrolled
enrolled in education

140


original histories


imputed histories

http://www.demographic-research.org



More intriguing information

1. Correlation Analysis of Financial Contagion: What One Should Know Before Running a Test
2. THE CO-EVOLUTION OF MATTER AND CONSCIOUSNESS1
3. Large-N and Large-T Properties of Panel Data Estimators and the Hausman Test
4. What Drives the Productive Efficiency of a Firm?: The Importance of Industry, Location, R&D, and Size
5. Has Competition in the Japanese Banking Sector Improved?
6. INTERACTION EFFECTS OF PROMOTION, RESEARCH, AND PRICE SUPPORT PROGRAMS FOR U.S. COTTON
7. The name is absent
8. Telecommuting and environmental policy - lessons from the Ecommute program
9. An Estimated DSGE Model of the Indian Economy.
10. Modelling the health related benefits of environmental policies - a CGE analysis for the eu countries with gem-e3