Decentralised CATI Versus Paper and Pencil Interviewing: Effects on the Results in the Swedish Labour Force Surveys
Lars R. Bergman, Karl-Erik Kristiansson, Anita Olofsson, and Margaretha Säfström
The results of decentralised computer assisted telephone interviewing (DCATI) and decentralised paper and pencil interviewing (DPAPI) were compared in a large methods study within the Swedish Labour Force Surveys. In the main study (n=103,161 and 23,420 for DPAPI and DCATI, respectively) key estimates were compared for the two methods. Very similar results were obtained except for (a) the category weak attachment to the labour market for which a 2% difference was found, believed to have been caused by a different layout of the response alternatives in DCATI and DPAPI, and (b) the percentage working 40 hours for which also a 2% method difference was found. Results from a reinterview study (n=1,800 for both DPAPI and DCATI) indicated that the two methods had produced data of very similar quality. In a coding study (n=4,127), the quality of the information provided by the interviewers on occupation and industry in DCATI was compared with previously obtained information using DPAPI. Both methods seemed to have produced almost equivalent information.
CATI; CAPI; computer assisted interviewing; telephone interviewing; data quality; measurement error; labour force survey.