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

Demographic Research: Volume 21, Article 6

Table 1: Post-secondary degrees attained by time of interview: degree
combinations^a. Women born in West Germany in 1964, interviews
in 1998/ 99

Source: GLHS, cohort 1964

(sample used for empirical analyses, excluding cases with missing values for education)

Notes : ^a ‘Basic vocational degree’ is our translation for 'Ausbildungsabschluβ,' ‘master craftswoman’s/ technician’s degree’
translates as ‘Meister’ or ‘Techniker,’ and ‘university or college degree’ refers to ‘Hochschulabschluft or
'Fachhochschulabschluβ/ A master craftswoman’s degree is a higher level vocational degree giving permission to train
apprentices and, at least in the past, also necessary to head an enterprise specializing in one’s vocation.

^b basic vocational or master craftswoman’s/ technician’s degree

As can be seen in Table 1, it appears that a substantial proportion of women in
cohort 1964 hold multiple degrees at the same level. Therefore, for the western German
context, it might make a difference whether in a survey, respondents are asked for the
first date or the last date they received a degree at a given level. If, as we suspect, gaps
between training spells are small, then it would be better to ask for the last date the
respondent received her highest degree. Our method of imputation would then consider
respondents to be continuously enrolled until the last time they received a degree at
their highest level. If gaps between training spells are small compared to the length of
the last spell, the extent of misrepresentation would be smaller using the last than the
first date the respondent received a degree at her highest level. To investigate whether,
for the purpose of imputing education histories, it would make a difference for survey
questionnaires to ask for the first or the last date respondents received a degree at their
highest level, separate imputations simulating each case are conducted for the empirical
analyses.

Another way to ensure that we get the last date a respondent received a degree is to
provide more differentiated degree levels in the survey questionnaire. In a simple list of
degree categories, a basic vocational degree and a master craftswoman’s degree, for
example, would both just be considered to be a vocational degree. If, however, the
survey explicitly differentiates between basic vocational degrees and master

http://www.demographic-research.org

<<   <   8   9   10   11   12   13   14   >   >>