Data archives (data and codebooks):
Principal investigator(s): K. Aarts (University of Twente, the Netherlands), H. van der Kolk (University of Twente, the Netherlands), M. Kamp (University of Twente, the Netherlands)
Fieldwork organization: Institute of Applied Social Research (ITS) at the University of Nijmegen
Design: This study consisted of 5 major parts: the first and second wave of interviews, a self-completion questionnaire (these three parts are called ‘the main study’), a non-response project, and a stimulus-effect project.The first wave started on March 30 and ended on May 5 (one day before the parliamentary elections). The second wave was held after the election date of May 6, namely from May 11 till July 4. At the end of the second wave a self-completion questionnaire was handed over to the respondents. They were asked to fill in the questionnaire and return it to the fieldwork organization.In the non-response project non-respondents were urged once more to grant an interview. As many as possible of those refusing to be interviewed face tot face were asked why they would not cooperate and were asked questions on political interest, efficacy and voting behaviour. These questions were identical to the questions asked to the respondents of the first wave. The stimulus-effect project was conducted to facilitate the study of interview participation effects. A random sample of persons not included in the main sample was contacted by telephone and asked about their willingness to grant a face to face interview and their voting behaviour. They were also asked some questions about political interest and political efficacy in order to allow comparisons between the main study, non-response project and stimulus-effect project.
Fieldwork dates: The fieldwork was done from March 30 till May 5, 1998 and from May 11 till July 4, 1998. The election date was May 6, 1998. Also after the Parliamentary election of 2002 on May 15, interviews were held for the panel study.
Sampling frame: This study contained a multi-stage sampling procedure. The DPES 1998 was based on a sample of persons rather than addresses. In the first stage municipalities were selected. The desired net sample size had been set at 2.000. The minimum number of addresses to be sampled for each municipality had been set to 1 percent of the net sample size (i.e. 20 interviews). Only 8 municipalities were self-selective. 83 additional municipalities had to be selected. This was done in a stratification procedure. The strata were determined by region and degree of urbanization.In order to select individuals, cooperating municipalities were asked to draw a sample from the population register. The original sample consists of 4.564 units, including 4.331 individuals and 233 addresses (addresses from mail-delivery registers, in the case when municipalities decided not to cooperate). Some of the addresses were not inhabited or did not include enfranchised voters. The gross sample consisted of 4.207 eligible voters.The main reason for selecting the population registers as the sample frame rather than the mail delivery register, was the problem of non-response in the DPES. The personalized approach of respondents associated with this sample frame was expected to reduce non-response. Also, information about age and sex of both respondents and non-respondents would help to analyze non-response.In 2002, post-election interviews were conducted with respondents from 1998. An attempt was made to re-interview as many as possible of the 1.814 respondents from the post-election wave of the 1998 DPES.
Response: The response in the first wave was 49.9 percent, which means that 2.101 of respondents completed the interview (gross sample 4.207). 1.814 respondents completed the interview in the second wave, so the response was 86.3 percent of the first wave. The response rate of the self-completion questionnaire was 66.1 percent (1.199 of the 1.814 respondents filled in and returned the questionnaire). 926 initial non-respondents (total non-response in the first wave was 2.106) were interviewed in the non-response project. In the panel study there was a response of 16,4 percent of the first wave and a response of 38 percent of the second wave (690 respondents completed the interview in 2002). Similar to 1998, the respondents were also asked to complete and return a written questionnaire concerning media usage; 637 respondents of the 690 did so.
Panel study: This election study did include a panel study. The interviews for this panel study were held in 2002 after the Parliamentary election date of May 15.