Data archive (data and codebook)

Main study: at DANS (P1628); at ICPSR (...) 

Principal investigator(s): G.A. Irwin (University of Leiden, the Netherlands), J.J.M. Holsteyn (University of Leiden, the Netherlands)J.M. den Ridder (University of Leiden, the Netherlands)

Fieldwork organization: NIPO (now TNS-NIPO, Amsterdam)

Design: This study consisted three waves. The first and second wave of interviews were held in 2002. In the pre-election wave face-to-face interviews were conducted. The post-election wave included some small scale experimental data collection methods. The third wave of interviews was held in 2003 after the Parliamentary elections of 2003. Because the election of 22 January 2003 followed so quickly after that of 15 May 2002, the DPES 2002 appointment was extended to include the DPES 2003.

Fieldwork dates: The fieldwork was done from April 18 till May 14, 2002 and from May 16 till June 27, 2002. The election date was May 15, 2002. The interviews in 2003 were held from January 23 till March 3. The Parliamentary election of 2003 was held on January 22.

Sampling frame: This study contained a multi-stage sampling procedure. The first step in selection of the sample in 2002 was to determine those municipalities in which interviews would be held. The list of municipalities was downloaded from the Statline website of the Statistics Netherlands on December 13, 2001. The list was then sorted according to region and degree of urbanization. The selection of municipalities was done by systematic random sampling. Interviews were held in a total of 90 municipalities. In these municipalities TNS-NIPO generated a random sample of 8893 telephone numbers. Some numbers were not related to a household, other numbers were not usable. Among those who were contacted (n=5791), 3541 refused to participate and 2250 were to be interviewed. In the pre-election phase 1907 interviews were completed. In order to guarantee continuity with previous election studies, most of the respondents (n=1200) were re-interviewed face-to-face in the post-election wave. A total of 374 respondents made their own choice, some of these respondents received a written questionnaire, some received a diskette containing the questionnaire, some hade chosen a face-to-face interview and some had chosen to fill out the questionnaire via an online connection to TNS-NIPO. So 1.574 respondents participated in the post-election wave.When the Dutch government fell in October 2003, it was announced that a new election would be held on January 22, 2003. Because this came soon after the previous election, there were numerous problems for the design and organization of a new DPES. For many reasons it was decided to collect data only after the election had been held. An attempt was made to interview all respondents who had participated in the 2002 post-election wave (n=1.574). As the time between the elections was relatively short and interests in politics had generally remained high or maybe even increased, it was possible to re-interview 1.271 respondents.In addition to this panel study, a fresh sample was drawn in the same manner as in 2002. The number of municipalities was 63. The total number of respondents was 1.287.

Response: The response in the first wave was 32.9 percent, which means that 1.907 of respondents completed the interview (gross sample 5.791). 1.574 respondents completed an interview in the second wave, so the response was 82.5 percent of the first wave. The response rate in the third wave (panel study) was 66.6 percent compared to the first wave and 80.7 percent compared to the second wave.

Panel study: This election study did not include a separate panel study. There was a panel element in this study, namely that respondents from the pre- and post-election wave of 2002 were re-interviewed in 2003.