Research programCentERdata is very experienced in the field of data collection and panel research. We use this experience to design and implement research in the field of methodology and model development. This is how we make a contribution to science and keep our expertise up to date. Ongoing research:Part of the questions in the LISS panel originates from large (inter)national surveys. Comparing these data with the data from the original survey can reveal mode and sample effects. In addition, the CentERpanel data and LISS panel data are compared in order to study the effects of differences in panel design. Panel conditioning When people participate in a panel, they may be susceptible to panel conditioning. We wish to investigate to what degree this takes place. Data from established long-time CentERdata panel members are compared with data from the new LISS panel. Data dissemination CentERdata wishes to give a new impetus to the quality of data dissemination and data adaptation for secondary analysis. We are developing software that is to meet the latest Data Documentation Initiative (DDI) standards. Methods and techniques of online questionnaires Designing online questionnaires requires choices to be made on which interface, questions, or answer categories should be used. Each question can be displayed on the computer screen in different ways. We investigate the influence of different methods and techniques on data quality and interpretation, to find out whether a given answer category is or is not appropriate for a particular question. Some questions are better suited to closed answers, whereas others favour an open format. Hard-to-reach groups Certain groups of respondents are virtually always underrepresented in survey research. The latest findings in the literature have been integrated into the LISS panel design, in order to achieve the highest possible response rates in all groups. Further research should establish to what degree this has been successful. Non-response bias in panels Non-response bias occurs when specific groups refuse to participate in the panel. In addition, panel members may refuse to complete a questionnaire or not complete a questionnaire in its entirety. We attempt to understand what are the different kinds of bias involved in non-response. In addition, we also investigate the questions that cause respondent dropout. Model development Our research into model development focuses on the theme of the labour market and involves so-called structural micro-simulation models. An example of such research is the Mimosa model. This model focuses on individuals’ participation decision, taking into account individual characteristics, household characteristics, legislation and regulations, and market conditions.
In the framework of model development, a research project that is currently ongoing looks into volunteer caregiving. Using a structural model, we are investigating adults’ decision-making process to provide care to their parents, which includes the modelling of decisions between siblings. Elements that play a role in informal caregiving include the following: the parents’ state of health, travelling distance between parents and children, the number of brothers and sisters, the wage base, and the household situation of adult children. We need to understand the adults’ decision-making processes because, in consequence of population ageing in society, governments intend to stimulate both labour participation and volunteer caregiving. The model allows the evaluation of political measures. This research project into volunteer caregiving is carried out in collaboration with Peter Kooreman (Tilburg University) and Matthijs Kalmijn (Tilburg University). |
