-20- 
III. DATA AND MEASUREMENT 
3.1. The Life History Study 
As part of the German Life History study, the Max Planck Institute for Human Development 
and Education (Berlin; principal investigator: Karl Ulrich Mayer) in cooperation with Infratest 
(Munich) collected life course data of 200421 men and women born between 1954-56 and 1959-61 
from November 1988 until October 1989 (Brückner & Mayer, 1995). A computer-assisted telephone 
interview (CATI) was designed to facilitate and support retrospective data collection. The ques¬ 
tionnaire included questions on migration history, education, job career, family of origin and own 
family, health, as well as voting behavior and standard demographic variables (see Figure 1 for an 
overview over most of the content and the structure of the questionnaire). 
The CATI technology offers opportunities to incorporate simple consistency checks, 
personalized questions, highlighting of important stimuli, and monitoring the ongoing interview 
process. — Considerable effort went into programming, interviewer training, and supervision. A two- 
day training session with mock interviews provided interviewers with a basic familiarity with the 
questionnaire and measurement concepts. In addition, each interviewer obtained printed material 
pointing out critical parts of the schedule, especially ramifications of the German educational system 
with related concepts and classification problems in the migration history. 
Respondents were called during the evening (between 6 and 9 pm). On the average, 8 
interviewers worked together in one shift. For the first 7 months of the survey, two supervisors were 
responsible for scheduling and monitoring of interviewers. In the remaining 5 months, one supervisor 
was felt to suffice. 
All completed interviews were reviewed in a two-stage editing process. Extensive consistency 
checks between and among the available life domains were made, using tape recordings (available for 
about two thirds of all cases), as well as call-backs to respondents for clarifying remaining inconsisten- 
The actual number of completed interviews was 2008. Four interviews were conducted 
by two senior researchers and two supervisors and are therefore excluded from the analysis which 
is based accordingly on 2004 cases and 24 interviewers. 
For example, answers given earlier in the interview would be used to ask questions 
which incorporated the specific context such as "And when did you start working as a self- 
employed carpenter?"