Generations and Gender Survey Netherlands Wave 1, Wave 2 & Wave 3

WAVE 1 MODE OF DATA COLLECTION Method: Face-to-Face (personal interview) + Self-administered questionnaire. Technique: Computer-Assisted (CAPI) and Paper and Pencil (PAPI), respectively.

WAVE 2 MODE OF DATA COLLECTION Method: Interview questionnaire (CAPI / CATI / CAWI) + Self-completion questionnaire (PAPI / CAWI) Technique:

• PAPI: Paper and Pencil (self-completion questionnaire only)
• CAPI: 94.8%
• Computer-Assisted Telephone Interviewing (CATI): 3.3%
• Computer-Assisted Web Interviewing (CAWI): 2% Note that the CATI and CAWI options were offered only at a later stage of data collection, when sample members who had previously refused to participate or had not been reached were re-contacted.

WAVE 3 MODE OF DATA COLLECTION Method: Mixed method: CAPI, CATI, CAWI. Different compared to wave 1 and 2, data were mainly gathered by Web Interviewing. CATI and CAPI were administered if respondents chose for these options. The majority of the data was collected by Web (55 %) and about 27 % was gathered via the telephone (CATI).

WAVE 1 ACTIONS

1. Dealing with nonresponse 1.1 Screening: Yes. Checks were carried out on the correctness of address information. If necessary, a special unit of interviewers personally visited addresses to find out who lived there. 1.2 Refusal conversion: Interviewers who managed to convince people who had initially refused, to take part after all, received a bonus. The initial plan was to re-contact soft-refusers only. However, given that only a small proportion of the refusals were categorized as soft (around four percent), the entire group of refusers, both soft and hard, were re-contacted. Addressees with a telephone number were re-contacted by a call center, whereas expert interviewers re-contacted addressees without a telephone number. Starting in May 2003 travel cheques were raffled as an additional incentive. In addition, an additional incentive of €15 was offered to initial refusers. 1.3 Incentives: Yes. At the start of the field work, respondents were given a gift voucher worth €10,- at the end of the interview, irrespective of whether they would complete the self-completion questionnaire. Starting in May 2003, travel cheques were raffled as an additional incentive. Potential respondents were informed about this in the introductory letter. Another additional incentive which was introduced in May 2003, was to offer potential respondents who were not willing to participate on the basis of a €10,- incentive, €15,- extra (so they would receive €25,- in total). The extra €15,- was not a particularly effective incentive. Less than one percent of sample member were converted.

2. Tracking of sampled units 2.1 Respondent contact information: Yes, respondent's address and telephone number (home, mobile, work) were collected. The respondent was not explicitly asked to provide an e-mail address. 2.2 Other contact information: Yes, contact details (address and telephone number [home, mobile, work]) of a proxy (such as parents, friend) were collected. The person was not explicitly asked to provide an e-mail address. 2.3 Cards: A "thank-you-letter", "change-of-address-notification" card (with paid postage fee, included in all the mailings as a standard part of the letter), and birthday/holiday cards were sent to all the respondents (planned frequency of contacts with the respondents between the two waves: at least annually). Furthermore, an on-line form for notification of change of contact details were set up, a brochure with results of Wave 1 was sent, and there were personal contacts such as telephone calls or visits. 2.4 Additional surveys: Yes, 12 mini-panels; qualitative interviews. For an overview of the topics see http://www.nkps.nl/NKPSEN/MiniPanels.htm 2.5 Administrative records: Yes. Respondents were encouraged to report a residential relocation. Change-of-address cards were included in mailings, and respondents were also informed about the option of entering new address information in a form constructed for that purpose on the NKPS-website. Checks were also carried out with the help of municipal population registers.

WAVE 2 ACTIONS

1. Dealing with nonresponse 1.1 Screening: Yes. In the period between the first and second wave, respondents were encouraged to report a residential relocation (see above documentation about Wave 1 "2.5 Administrative records"). In preparation of Wave 2 fieldwork all addresses, all addresses were validated by Cendris, a daughter company of Dutch Mail (TPG Post) that holds a database of more than 7.3 million private households in the Netherlands. During the course of fieldwork, additional efforts were made to track down the correct address via Cendris. 1.2 Refusal conversion: Three re-contacting phases were carried out. After failure of first attempts, 'soft' refusals were approached by a different interviewer in the re-contacting phase-after a slightly different introductory letter had been sent; 'hard' refusals were phoned by specially trained CATI interviewers. The third phase concerned proposal of alternative methods. Additional incentives were offered to refusers (The average value of the incentives given to all respondents in the re-contacting phase was 20 Euros). 1.3 Incentives: Anchors who agreed to take part after the first contacting phase were given 10 Euros (option of a giftcard or donating). In the re-contacting phase respondents were offered on average 20 Euros. Respondents were offered additional incentives if they did not return the self-completion questionnaire in time.

2. Tracking of sampled units 2.1. Respondent contact information: Between wave 1 and 2 respondents were encouraged to report address changes. Before the start of wave 2, all addresses were validated by Cendris through their database of 7.3 million private households in the Netherlands. 2.2. Other contact information: DK 2.3. Cards: Interviewers filled in contact forms for up to ten initial visits to each address, which was afterwards entered into the CAPI laptop. The interviewers had access to respondents' contact information with each new phase of contacting the respondents. 2.4. Additional surveys: Yes, 12 mini-panels; qualitative interviews. For an overview of the topics see http://www.nkps.nl/NKPSEN/MiniPanels.htm. 2.5. Administrative records: Yes. Respondents were encouraged to report a residential relocation. Change-of-address cards were included in mailings, and respondents were also informed about the option of entering new address information in a form constructed for that purpose on the NKPS-website. Checks were also carried out with the help of municipal population registers. 93% of the respondents accepted to link questionnaire information to registry data.

WAVE 3 ACTIONS

1. Dealing with nonresponse 1.1 Screening: In the period between the second and third wave, respondents were encouraged to report a residential relocation. In preparation of Wave 3 fieldwork all addresses, all addresses were validated by Cendris, a daughter company of Dutch Mail (TPG Post) that holds a database of more than 7.3 million private households in the Netherlands. During the course of fieldwork, additional efforts were made to track down the correct address via Cendris (NKPS Codebook Wave 3, p17). 1.2 Refusal conversion: After failure of first attempts, 'soft' refusals were approached by home visit of CAPI interviewer and 'hard' refusals were phoned by specially trained CATI interviewers. If respondents agreed, they could also choose the survey mode: CAWI, CATI or CAPI (NKPS Codebook Wave 3, p.22). 1.3 Incentives: Respondents were presented with the option of receiving a gift certificate or of having the money donated to charity. Anchors who agreed to participate after the first contact phase were give €10. Interviewers were give the freedom to offer some respondents more than others, up to €25 during the re-contacting phase. Partners did not receive an incentive when they participated in the first contact phase, but were offered €5 in the re-contacting phase.

2. Tracking of sampled units 2.1 Respondent contact information: Between wave 2 and 3 respondents were encouraged to report address changes. Before the start of wave 3, all addresses were validated by Cendris through their database of 7.3 million private households in the Netherlands (NKPS Codebook Wave 3, p17). 2.2 Other contact information: For most respondents phone number were also known ((NKPS Codebook Wave 3, p21). 2.3 Cards: Interviewers filled in contact forms for up to ten initial visits to each address, which was afterwards entered into the CAPI laptop. The interviewers had access to respondents' contact information with each new phase of contacting the respondents (NKPS Codebook Wave 3, p23). 2.5 Administrative records: Respondents were encouraged to report a residential relocation. Change-of-address cards were included in mailings, and respondents were also informed about the option of entering new address information in a form constructed for that purpose on the NKPS-website (NKPS Codebook Wave 3, p17).

WAVE 1 DATA COLLECTION

1. Interviewers 1.1 Total number of interviewers: 160. 1.2 Number of interviewers in the field: 5 during pilot study, 160 during main study. 1.3 Network organization: Centralized. 1.4 Working arrangement of interviewers: Professional interviewers, many working free lance. Some worked full-time, others worked part-time. 1.5 Payment of interviewers: The interviewers were paid by interview. The income per interview included remuneration for the time it takes to contact respondents, travel to their home, and complete all the logistics of the interview. Interviewers received separate remuneration for attending training sessions (which was mandatory). In May 2003, a bonus scheme was introduced for interviewers. Under the scheme, interviewers who succeed in conducting five or more interviews a week received a financial bonus. Interviewers who managed to convince people who had initially refused to take part after all, also received a bonus. Interviewers' conversion efforts were restricted to sample members without a known telephone number.

2. Interviewer training: 2.1 General interviewing: Yes, basic training in skills such as how to contact respondents, how to ask questions, and how to deal with problem situations. 2.2 Survey specific: Yes, during the NKPS/GGS-specific training, the interviewers received background information on the study, and were given instructions on topics specific to the NKPS, such as how to select Anchors, deal with multiple households, obtain Survey Alter addresses, classify occupations, and use the built-in postal code table. Part of the training involved practice with the CAPI questionnaire. 2.3 Length: General training: two days; NKPS/GGS-specific training: one day. 2.4 Control of performance: Yes. To monitor the quality of the interviewers' work, members of the fieldwork team listened to audio tapes of their interviews. Each interviewer was required to record the first three interviews on tape, as well as another ten percent during the remainder of the fieldwork (the computer prompted when this had to be done). In addition, a Response Analysis Survey (RAS) was carried out by telephone among ten percent of the addresses. Both respondents and non-respondents were presented with a short set of questions designed to assess selectivity in response. Finally, if a questionnaire had incomplete name and address information on an Anchor, the interviewer was requested to track down the missing information. In the case of Survey Alters, where possible, missing address details were tracked down by the fieldwork team with the aid of public sources of information (telephone directories, postal code table). 2.5 Interviewer survey: Yes. After each interview, the interviewer was asked to provide the following information: general atmosphere during interview, problems of understanding, assistance needed by answering, difficulties with answer categories, digressions in answering questions, interest in interview, suspicion about the research, honesty of answering, special conditions during the interview. Refresher sessions were organized during the course of the fieldwork, allowing the interviewers to address certain questions or problems they had encountered to the fieldwork coordinator. Finally, a self-completion questionnaire (designed by Joop Hox and Edith de Leeuw, experts in interviewer effects) was administered regarding their operating procedures and attitudes.

3. Contact protocols 3.1 Advance letter: Sample members received an informative and inviting introductory letter describing the purpose of the survey. The letter was personalized, if possible. A brochure was included with the introductory letter, providing information about the background to the survey, how the survey was being conducted, the importance of the respondents' participation, what the results would be used for, privacy protection, and the organization and individuals involved in the survey. Addressees were informed that the interviews had an average duration of 75 minutes. There was no mentioning of data collection among family members, because addressees might become apprehensive and decline participation. The panel nature of the study was also not emphasized, neither the inclusion of a self-completion questionnaire. The letter talked about families in a balanced way, to avoid running the risk that only people from closely-knit families would take part in the study; both positive and negative aspects of families were touched upon. Addressees were informed that they would receive €10,- if they participated in the interview. For additional information, addressees were referred to the NKPS website. 3.2 Cold contacts: All addressees were sent an introductory letter before the first contact attempt (no cold contacts in the sample). For those who had not been reached after the maximum number of contact attempts (ten phone calls or three home visits), a new attempt was made in May 2003. If a telephone number was available, a call center would re-contact the addressee, whereas interviewers re-contacted addresses without a telephone number.
   3.3 Scheduling / scattering: Yes. For addresses with telephone numbers, at least ten attempts at different times of day and on different days of the week were made. For addresses without telephone numbers, at least three attempts at different times of day and on different days of the week were made. The fieldwork bureau monitored the timing and number of contact attempts. 3.4 Contact history: Yes. For each contact attempt the interviewer had to record the time the attempt was made, how contact was made, and the result of the contact. 3.5 Min number of contacts: For addresses with telephone numbers, at least ten attempts were made. For addresses without telephone numbers, at least three attempts were made. 3.6 Max number of contacts: NA.

4. Questionnaire localization 4.1 Validation: NA. 4.2 Pre-test: A pilot was carried out in April 2002 (55 respondents), discovering some glitches in question routing. Furthermore, interviewers reported the need for introductory sentences, elaborations. 4.3 Length of interview and additional information about the data collection situation: Average length of interview: 74 minutes. Ideally, the interview was to be conducted with no other persons present. If this situation was not possible (e.g. the presence of young children), interviewers reported the presence of others (by filling in a pre-structured questionnaire on the interview situation). They reported who was present (e.g. partner, son/daughter), whether the other was present during part or during the entire interview. Sensitive questions (e.g. attitudes, feelings of loneliness) were addressed via the self-completion questionnaire. During the course of fieldwork, it became apparent that people under the age of 30 were under-represented. Starting in September 2003, call center employees and interviewers who contacted addresses specifically requested to speak to someone over the age of 18 and under the age of 30.

WAVE 2 DATA COLLECTION

1. Interviewers 1.1. Total number of interviewers: 153 1.2. Number of interviewers in the field: 150. During the fieldwork, three interviewers left; two chose to do so, and one was requested to leave by GfK PS. Another 17 interviewers were trained in telephone interviewing. One left on own request. 1.3. Network organization: DK 1.4. Working arrangement of interviewers: DK 1.5. Payment of interviewers: DK

2. Interviewer training: 2.1 General interviewing: Organized by GfK PS, Dufec, and the NIDI Fieldwork Centre. Top priorities: (1) achieving a high and non-selective response, (2) obtaining survey Alter addresses, (3) receiving Anchor's permission to be contacted for participation in possible third wave, (4) receiving Anchor's permission to link survey to registry data. General interview training included basic training in skills such as how to contact respondents, how to minimize non-response. In addition to the training sessions organized at the start of the fieldwork, refresher days were organized for the interviewers. During these sessions, additional instructions were given and there were opportunities to ask questions and exchange experiences. Furthermore, sound recordings of interviews were played in order to address specific quality aspects of interviewing. 2.2 Survey specific: Background information about the NKPS and the rational underlying the interview questionnaire was provided. Also, a special session was devoted to practicing the routines in the interview questionnaire, including the built-in tools (i.e. postal code table, and occupational classification table). 2.3 Length: 10 days with 2 half-day sessions, 2 days for telephone interviewers, refresher days 2.4 Control of performance: (1) Interviewers were monitored on the basis of initial contact, response rates, and recording of the interviews (if permitted by the respondent). (2) 1500 control questionnaires were sent to anchors to reveal potential cases of fraud. (3) Checks were carried out by NIDI fieldwork centre were municipality's and interviewers' address information differed. (4) Progress was monitored at distance through an information processing package. 2.5 Interviewer survey: Yes, The interviewer evaluated the course of each interview after it had been completed. Questions were answered concerning the setting of the interview, such as the presence of others and the general atmosphere. The Anchor's performance was also rated, using the evaluation form of the interview questionnaire.

3. Contact protocols 3.1 Advance letter: Personalized introductory letter and a brochure were sent to the respondent two days before the Anchor was contacted. The letter informed about the toll-free telephone number of the NKPS Helpdesk and the NKPS e-mail address. Furthermore, addressees were informed about the duration of the interview (about one hour) and that they would receive €10,- if they participated in the interview. They were also told that after all interviews had been conducted, three travel vouchers among all participants (1 x € 500 and 2 x € 250) were to be put up for raffle. For additional information, addressees were referred to the NKPS website and to the brochure. 3.2 Cold contacts: All addressees were sent an introductory letter before the first contact attempt (no cold contacts in the sample). For those who had not been reached or who has refused to participate after the maximum number of contact attempts (ten phone calls or three home visits), a new attempt was made. If also this re-contacting attempt failed, alternative survey methods were offered: CATI and CAWI. Respondents willing to participate via CATI were contacted by the call center for the completion of the questionnaire. Respondents willing to participate via CAWI were sent a letter stating the website address and their personal access code. Respondents who subsequently failed to complete the questionnaire were sent a letter of reminder. 3.3 Scheduling / scattering: Yes. Phase 1: All participants were asked to participate in the CAPI questionnaire through introductory letter, followed by a phone call. Phase 2: Soft refusals: Same procedure as in phase 1 but with different interviewer. Hard refusals: Contacted for CATI questionnaire. Phase 3: CATI and CAWI questionnaires were offered. 3.4 Contact history: Yes. For each contact attempt the interviewer had to record the time the attempt was made, how contact was made, and the result of the contact. 3.5 Min number of contacts: 10 initial contact by telephone, 3 home visits 3.6 Max number of contacts: NA.

4. Questionnaire localization 4.1. Validation: A pilot was carried out in the months from May until July 2006 (200 respondents). The purpose of the pilot was to test the different data collection methods and to follow all related procedures. 4.2. Pre-test: YES.

   • 141 Dutch respondents (CAPI, CATI, CAWI)
   • 35 Dutch-speaking migrants (CAPI)
   • 24 non-Dutch speaking migrants (CAPI) 4.3 Length of interview and additional information about the data collection situation: First approach: Respondents were encouraged to take the CAPI questionnaire. First alternative: CATI Second alternative: CAWI

CAPI procedure:

1. Interview at the respondent's home
2. Anchor receives self-completion questionnaire, which is either returned by Anchor or picked up by interviewer

CATI procedure: Respondents were interviewed by a call centre

CAWI procedure: Respondents received letter with website and login data

Average duration: 60.1 minutes (60.1 minutes via CAPI, 63.2 via CATI, 51.6 via CAWI). To ensure that the questionnaires for CAPI and CATI were identical (apart from a few textual adjustments) the telephone team also conducted the interview with the aid of CAPI laptops. Apart from textual differences linked to the method of data collection, the content of the CAWI questionnaire differs slightly from that of the CAPI/CATI questionnaire, namely: differences in the questions (e.g., different routing for questions on income), and differences resulting from procedural differences (e.g., the CAWI questionnaire does not include built-in controls).

WAVE 3 DATA COLLECTION

1. Interviewers 1.1. Total number of interviewers: CAPI: 69 with 2 supervisors; CATI: 39 interviewers with 8 supervisors 1.2. Number of interviewers in the field: 4 CAPI interviewers and 2 CATI interviewers dropped out1.3. Network organization: DK 1.4. Working arrangement of interviewers: DK 1.5. Payment of interviewers: DK

2. Interviewer training: 2.1 General interviewing: Suitable candidates (preferably those who already were involved in collection of previous waves) were invited to attend a special training session provided by NKPS team members in collaboration with GfK PS. Only interviewers who had participated in the training session were allowed to conduct interviews. 2.2 Survey Specific: The structure and content of the training sessions were designed by GfK PS in cooperation with the NKPS team members. The program consisted of a half-day session in which background information on the NKPS was provided and the rational underlying the interview questionnaire was explained. The four top-priorities were: (1) achieving a high and non-selective response, (2) obtaining permission to contact the current partners for participation as a Survey Alter, (3) receiving Anchor's permission to be contacted for participation in a possible fourth wave, and (4) receiving Anchor's permission to link the survey data to registry data. 2.3 Length: 2 training sessions, 1 for CATI & CAPI interviewers and their supervisors and 1 for 6 other CATI interviewers. (NKPS Codebook Wave 3, p.20) 2.4 Control of performance: (1) Interviewers were monitored on the basis of speed, quality of interviewing, quality of initial contact, response rates, and two recording of the interviews. (2) 1000 control questionnaires were sent to anchors to reveal potential cases of fraud. 422 questionnaires were sent back and no cases of fraud were revealed. (3) Checks (i.e. personal visits to the call center) were carried out by the NKPS team and the NKPS project leader at NIDI and the GfK project leader had weekly contact via telephone to discuss procedures and progress. (4) Progress was monitored at distance through an information processing package.
   (NKPS Codebook Wave 3, p.26) 2.5 Interviewer survey: The interviewer evaluated the course of each interview after it had been completed. Questions were answered concerning the setting of the interview, such as the presence of others and the general atmosphere. The Anchor's performance was also rated, using the evaluation form of the interview questionnaire. Less than one percent of the interviews were rated as 'unpleasant' or 'very unpleasant' (NKPS Codebook Wave 3, p.28).

3. Contact protocols 3.1 Advance letter: The first contact was through an introductory letter in which respondents were informed about het upcoming third wave of het NKPS. The letter also directed respondents to the NKPS website, included a full-color brochure for further information about the study and referred to the toll-free telephone number of the NKPS Helpdesk, and the NKPS e-mail addresses. In addition, the following letters/emails were sent to the Anchors who wished to participate in the study through the Web: (1) Web letter of invitation, (2) Web reminder, and (3) Web second reminder. 3.2 Cold Contacts: Potential respondents were contacted in two phases. First, all addressees were sent an introductory letter before the first contact attempt, CAWI or CATI (no cold contacts in the sample). The interviewer contacted the respondents by telephone within two days of receipt of this letter (if their phone number was known). For respondents whose phone number is known, 10 contact attempts were made and three home visits if the phone number was unknown. In the second phase, re-contacting attempt, a distinction was made between 'hard' and 'soft' refusals. Hard refusals were approached by telephone by specially trained CATI interviewers, and soft refusals were personally approached by a CAPI interviewer. If the participant agreed to participate in the third wave, the participant could choose between the three interview modes: CAWI, CATI or CAPI. Respondents willing to participate via CATI or CAPI were contacted by the call center to set a date. Respondents willing to participate via CAWI were sent a letter stating the website address and their personal access code (NKPS Codebook Wave 3, p.21-22). 3.3 Scheduling / scattering: Phase 1: All participants were asked to participate in the third wave via CATI or the Web through an introductory letter, followed by a phone call. Phase 2: Re-contacting phase. 'Soft' refusals were personally approached by a CAPI interviewer; 'hard' refusals were phoned by specially trained CATI interviewers. 3.4 Contact history: For each contact attempt the interviewer had to record the day and time, the method, and the result of the initial contact ((NKPS Codebook Wave 3, p.24) 3.5 Min number of contacts: 10 initial contact by telephone, 3 home visits

4. Questionnaire localization 4.1 Validation: A pilot was carried out in the months in April and May 2010 (148 respondents). The purpose of the pilot was to test the material, the different data collection methods and to follow and evaluate all related procedures. test the different data collection methods and to follow all related procedures. 4.2 Pre-test: 35 CAPI; 41 CATI; 72 Web the CAWI questionnaire does not include built-in controls). 4.3 Characteristics of the data collection situation: First approach: Respondents received a letter with invitation for CAWI or CATI interview. alternative: CAPI CATI procedure: Respondents were interviewed by a call centre CAWI procedure: Respondents received letter with website and login data CAPI procedure: Interview at the respondent's home Average interview duration: 42 minutes (54 minutes via CAPI, 48 via CATI, 35 via CAWI). To ensure that the questionnaires for CAPI and CATI were identical (apart from a few textual adjustments) the telephone team also conducted the interview with the aid of CAPI laptops. Apart from textual differences linked to the method of data collection, the content of the CAWI questionnaire differs slightly from that of the CAPI/CATI questionnaire, namely: differences in the questions (e.g., different routing for questions on income), and differences resulting from procedural differences (e.g., the CAWI questionnaire does not include built-in controls). (NKPS Codebook Wave 3, p.8-28).