Online Program

Experiences with sequential mixed-mode interviewing in ESENER-3, a cross-national establishment survey on health and safety at work (308146)

Keywords: Mode effects; Cross-national establishment survey; response bias

The European Agency for Safety and Health at Work (EU-OSHA) conducted its third European Survey of Enterprises on New and Emerging Risks (ESENER-3) in 2019. The survey focuses on the management of occupational safety and health in practice, covering establishments with 5 or more employees from almost all sectors of activity in 33 European countries.

Though genuinely a telephone survey, in the 2nd and 3rd wave an additional online interview option was of-fered to respondents refusing to participate in the telephone interview. The main aim of the online option was to enhance response rates and minimize any selection bias. Overall, 2,166 (5% of the total 45,420) inter-views were completed online. Though all countries used the online option, the share of online interviews varies largely between countries.

The presentation will outline the practical challenges of implementing the online option and the changes made to ESENER-3 on base of the learnings from the previous wave (2014). It will then focus on 3 key re-search questions:

(1) Which effect did the CAWI interview have on response rates? Are the additional costs for the online interviews a useful investment?

(2) Who used the online version? Do establishments answering the online version systematically differ from respondents of the online interview, on aspects such as size, sector of activity or the function of the respondent?

(3) Do the CAWI interviews yield different results (mode effects) and thus potentially hamper cross-national comparability and comparability over time?

Answers to these questions will be based on both bivariate calculations and a series of multi-variate regres-sion models calculated with the survey data-set. While controlling for various structural aspects of the estab-lishments, the regression models analyse the effect of the mode on the answers to several key survey ques-tions.