(in German)

In the BIBB/BAuA Employment Survey (formerly BIBB-IAB Survey) interviews with employees about their working situation have been conducted since 1979. Until 1998/1999 the respondents were interviewed personally at their homes (CAPI; Computer-Assisted-Personal-Interviews). The survey method was changed to CATI (Computer-Assisted-Telephone-Interviews) in the fifth wave in 2005/2006.

In this project (by request of BAuA), in addition to the 20,036 interviews that were done with CATI, 2,000 respondents were interviewed using CAPI and 2,000 others using CAWI (Computer-Assisted-Web-based-Interviews) method. The first aim was to examine whether all three methods are suitable for an employment survey. The second aim was to investigate differences caused by the survey method.

The findings reveal that both CAPI and CATI are suitable for an employment survey. Nearly all psychometric criteria of the CAPI-survey showed better results than those of the CATI-survey. However, CATI is the more efficient method with regard to costs. The findings for the CAWI-survey indicate substantial limitations. On the one hand, there are significant deviations in the selection of the sample which cannot be compensated by weighting the data according to Microcensus. On the other hand, an interview length of more than 20 minutes is classified as critical for CAWI.

With regard to method-immanent differences between CAPI and CATI, different constructs (physical demands, psychological demands, psychovegetative complaints, musculoskeletal complaints) are analysed using confirmatory factor analyses and structural equation modeling. These constructs show comparable reliabilities and validities in both survey methods. The relationships between the constructs are shown to be stable and similar in all three survey methods.

In general, the findings reveal that CAPI and CATI are suitable for employment surveys but a comparison of different methods over time is not possible without complications.

Please download the complete report "Methodological experiment as part of the BIBB/BAuA Employment Survey 2011/2012: A comparison of CATI, CAPI, and CAWI" (in German).