10th Workshop on Labour Force Survey Methodology 2015, Prague
DATE 29 - 30. April 2015 TOPIC Data collection
VENUE Prague, Czech Republic
CERGE-EI – Thanks for representative ROOMS CERGE-EI, Prague is a joint workplace of The Center for Economic Research and Graduate Education (CERGE) of Charles University, Prague and the Economics Institute (EI) of the Academy of Sciences of the Czech Republic.
10th lfs methodology Workshop Each session consisted of presentations and discussion (the same time for each part) Presentations were focused on the main findings - usefulness for the other countries Content of presentation was limited only to highlight main points Details were presented in the paper
10th lfs methodology Workshop All information still available on the website (including presentations and papers) https://www.czso.cz/csu/czso/10th-workshop-on-labour-force-survey-methodology
Extra Topic – LFS in Enlargement countries Albania, Montenegro, Bosnia and Herzegovina, Kosovo Main points: an overview of the Labour Force Survey methodological information about the survey (sample, dissemination, data processing, sampling errors) further development (for example improving the sample frame, introduction of continuous survey, using CAPI for data collection) financial resources for the next years, maybe the main problem
Harmonisation of the measurement of employment and unemploymenT Presenters: Sammy Lauritsen (Euorostat), Kieran Walsh (ILO), Peter Beijron (Sweden), Vivian Meertens (Netherlands), Marc Plate and Melitta Fasching (Austria) Aim: Fully developed mode-independent questionnaire, with filters, routing and wording of the questions, giving code lists, implementing instructions and explanatory notes In this section the results from pre-testing LFS model questionnaires Main questions: One or more intro-modules? What are acceptable reasons to deviate?
USING MULTIPLE MODE DESIGN with focusing on CAWI Presenters: Thomas Koerner and Karen Blanke (Destatis), Pertti Taskinen (Finland), Peter Betts (GBR), Michael Frosch (Denmark), Sylvie Le Minez (INSEE) Main points: results from the ESSnet project Social Surveys mode effects – the limits of the CAWI (response, timeliness, problem with self-classification) different questionnaire design compared CAWI with CATI and CAPI Discussion: future of the CAWI for social surveys, and the possibility of multiple mode design in particular countries
Fieldwork Processes Presenters: Murat Şahabettinoglu (Turkstat), Hanka Peroutková (CZSO), Mateja Zgonec (Slovenia) Main points, tolls for improving fieldwork processes: movie „Guidelines for successful field interviewing“ to have own proprietary case management system cooperation of interviewers, regional coordinators, national coordinators and main methodologists simple redistribution of workload to interviewers and data gathering Discussion: quality checks, possibility what is the effective way of training interviewers, which data (analysis) provide to coordinators
Questionnaire design and testing, proxy answering Presenters: Judith Forster and Katrin Baumgartner (Austria), Francisco Hernandez and Antonio Rueda (Spain), Rita Ranaldi at al. (ISTAT) Main points, tolls for improving fieldwork process: proxy effect, impact on the quality of the results, proxy as a source for the item non response (HWUSUAL) the system of management of CATI, to ensure the high response in relation to short time after the reference week improving the allocation of foreign levels of education – the use of additional questionnaire Discussion: generally about the effectiveness of realization of particular tools to improve the quality of questionnaire design, proxy answering – to choose the right approach
Approaches to reduce errors in LFS Presenters: Alessandro Martini (ISTAT), Krister Nasen (Sweden), Olle Vilund (Norway) Main points, tolls for improving fieldwork process: dependent interviewing - saving cost? better quality? to estimate and quantify the size of the systematic measurement errors focusing on labour market groups of employed and unemployed have been misclassified as ”not in the labour force” – probably not only in the Swedish LFS integrated quality monitoring system minimize the source of non- sampling errors Discussion: ways to measure the non-sampling errors and to reduce this source of bias (for example the interviewer effect)
SEE YOU IN PRAGUE IN 2043 !!