Growing Up in Ireland Cohort '08 Wave 7 Pilot Report

Growing Up in Ireland

Growing Up in Ireland (GUI) is the national longitudinal study of children and young people, a joint project of the Department of Children, Disability and Equality (DCDE) and the Central Statistics Office (CSO).
GUI had its origins in the National Children’s Strategy: Our Children - Their Lives which set out goals and policy priorities in relation to children for the period 2000-2010. A collaborative model exists for the survey involving DCDE and the CSO working in partnership to deliver GUI.

Cohort '08 Wave 7 Pilot

Overall, the objective for the GUI Cohort '08 pilot was to gather information to inform the implementation and design of the instrument for the main survey collection. Pilot surveys are an important part of the survey design process and can provide valuable insights into the elements that did/did not work well in the survey instrument when used in the field. They allow us to not only collect quantitative data on how the instrument performed but also qualitative data via interviewer and respondent feedback. The pilot survey is not designed to give insights into the answers to specific questions or used for analysis purposes beyond informing the design of the instrument.
The specific objectives of the pilot can be broken down as follows:

• To test the data collection instrument in the field, obtain respondent and interviewer feedback and identify improvements required.
• To test the sample distribution process. It was essential that distribution and syncing of cases was managed effectively, to ensure the young person and parent cases were issued to separate interviewer tablets.
• To manage field collection within the 17-year window.
• Due to the need to collect data from both young person and parents, we sought to maximize data collection efficiency by testing the use of two tablets in the fieldwork collection. Using the two tablets was considered necessary to reduce the burden on the respondents and time spent in the household by the interviewer. CSO interviewers have never used multiple tablets for data collection, so identifying any challenges ahead of main collection was necessary.
• To evaluate communication with respondents and test approaches with under-represented groups.
• To test the weighing and measuring devices used in the physical measurement data collection process.

The Cohort '08 Wave 7 Pilot questionnaire design process began in early 2024. An iterative process of stakeholder engagement and requirements analysis was undertaken. Stakeholder engagement – led by the Department of Children, Disability and Equality (DCDE) - included consultations with policy makers across a range of government departments and public bodies, researchers, and non-governmental organisations.
Previous waves, as well as the older cohort, Cohort '98, were used as a baseline for the development of the new instrument. Existing themes that were deemed to be relevant during the requirements gathering phase were maintained. New and emergent data needs based on stakeholder feedback were incorporated. These requirements were reviewed and prioritised based on specific criteria:

1. The measure captures one of the key domains of the lives of 17-year-olds;
2. The topic is policy-relevant;
3. The measure is age- or stage-appropriate;
4. The measure captures a (potentially) dynamic process;
5. There is longitudinal consistency with earlier waves and/or cross-cohort consistency with Cohort '98 at 17;
6. The measure has worked well in previous research;
7. There is comparability with other international cohort studies;
8. The measures capture the potential diversity of experiences and pathways among the cohort and are inclusive of minority groups;
9. The topic being measured has sufficient prevalence and variance to be analysed;
10. The topic or question is engaging to participants;
11. The issue can realistically be measured in a valid way within the context of Growing Up in Ireland;
12. The topic is not covered in other data collection; and
13. There is a trade-off between respondent burden (and its potential impact on response rates and data quality) and the need to take account of the other principles listed.

Research was carried out by DCDE on data points, scales and topics included in other international longitudinal cohort studies to enable cross-country comparability. Meta data repositories such as the CSO’s own Colectica database were mined for standardised questions on a range of topics. Data requirements were reviewed against the meta data available in the CSO Administrative Data Centre (ADC). This was followed by an intensive review and refinement of the three questionnaires from Cohort '08 (Young Person, Parent and Principal; for more details, see below) in consultation with DCDE.
The GUI Cohort '08 Wave 7 Pilot data collection ran for a total of 12 weeks – from the 14th of October 2024 to the 30th of December 2024. Household interviews were conducted by face-to-face interviewers. A very small cohort were invited to participate via video call interviews, where an interviewer was not available locally. The school principal data collection was followed up in January 2025 via an online web-based instrument.

Key Themes

GUI seeks to measure pathways through childhood and on into young adulthood. Longitudinal consistency over data collection waves is therefore essential but must be balanced with age and stage appropriate measures. For Wave 7 the key domains were:

• Health
• Socio-emotional well-being
• Education/cognitive development
• Civic and economic engagement

While the design of the GUI study is focused around the young person, information from other respondents was also collected, specifically:

• Parents resident with the young person (included in Wave 7)
• School principals (included in Wave 7)
• Teachers (collected in previous waves, but not included in Wave 7)
• Parents not resident with the young person (“Non-resident parent”; collected in previous waves, but not included in Wave 7)
• Childcare Providers (collected in previous waves, but not included in Wave 7)

Cohort '08 Wave 7 Pilot Fieldwork

Sampling

This pilot represented the seventh wave of interviewing with Cohort '08, which was recruited into the project as a random sample from the Child Benefit Register in 2008. At Wave 7, the longitudinal pilot sample consisted of 351 young persons (including twins), from 343 families. For this pilot, a sub-sample of 252 young persons were selected, to provide sufficient cases to accurately test the instrument and distribution processes, within the timelines of the pilot. These cases were geographically spread across urban and rural areas and provided sufficiently varied household types to accurately test the instrument and other pilot objectives.

Data collection and Recruitment

Data collection was via interviewer administered computer aided personal interview (CAPI) with a self-complete computer aided self-interview (CASI) for sensitive questions. An initial contact letter was sent to both parents and young persons, outlining what the pilot survey was about; informing them that a CSO interviewer would call to their home; assuring them that their information would be kept confidential, anonymous, and secure; and providing a link to further information on the website and a phone number and email address for them to contact us. All documentation was provided in both Irish and English.
A letter was issued to households outside the reach of our field force inviting them to complete their survey via a video call. This was a small ‘proof of concept’ exercise, to determine the effectiveness of this mode, for cases who would otherwise not have been able to participate.
A letter of invitation was issued to schools, requesting the principal participate in a short computer aided web interview (CAWI). This letter included a unique access key to facilitate matching the school to the household. After 2-3 weeks, a telephone nudge was carried out to encourage participation, and collection was extended to secretaries and deputy principals.

Instruments

There were three instruments used in the pilot:

• Young Person (YP) Survey – administered to the child;
• Parent Survey - administered to the primary informant (Parent 1) and any other resident parent/guardian, or resident partner of the parent who acts in a parental capacity (Parent 2);
• Principal Survey – administered to the principal of the child's school.

Table 1 gives an overview of the topics included in each of the instruments.

Findings from the Cohort '08 Wave 7 Pilot

For the young person (YP) survey, the data collection resulted in 134 full and 17 partially completed interviews, giving an overall response rate of 59.9%. Among the partially completed interviews, we had 7 video calls completed. In these cases, participants did not answer the self-complete block (therefore this was recorded as partial).
For the parent survey, we had a total of 122 completed cases and 18 partial, giving an overall response rate of 55.6%. The completed cases include situations where both parents answered or only parent one.
During the household collection, 100 unique schools were provided by the young persons. As the school principal instrument was not collecting data on the young person personally, only one questionnaire was issued to a school in the case of multiple GUI respondents. The school principal data collection resulted in 57 returns (40 completed and 17 partial), with an overall response rate of 57.0%.

Sample Distribution Process

• The sample distribution process worked as intended – the interviewer received the proportion of sampled households. Syncing parent and young person cases onto the separate tablets was successful and was monitored during training and field collection.
• Interviewers had success in interviewing within the 17-year window.
• The hundred schools for whom contact details were obtained were successfully distributed for a CAWI collection after the end of the main fieldwork.

Materials

• All materials were translated into English and Irish and well received by households. There were no queries raised related to data privacy.

Interview Length

• In the YP pilot survey, the median length of a completed interview was 1 hour 26 minutes. In the parent survey, the median length of a completed interview was 1 hour
  and 5 minutes for parent one and 28 minutes for parent two. In the principal survey, the median length of a completed survey was 24 minutes.
• The survey length increased considerably for twin households, taking up to 4.45 hours.
• Interviewers found the interview to be overly long, with the sections on education and food habits particularly long or exhausting to work through.

Communication with Respondents

• Pilot respondents were initially contacted through letters, informing them that an interviewer would contact them soon to arrange an interview date. Interviewers did not report any issues with initially contacting respondents via phone calls, and respondents were eager to continue their participating in the GUI survey.
• The gifts for the young person (water bottle and rubber duck) were well received, with feedback that they should perhaps be providing a gift to parent one as they were most involved in scheduling the interview and engaging with the interviewers.
• Principals were somewhat reluctant to complete the school instrument. Opening the collection to secretaries and deputy principals was successful.
• The proof of concept of using video calls as a mode of collection showed promise and the CSO is working towards incorporating this mode into our data collection model. There are challenges around the sensitive block which need further consideration.

Interview Content

• Interviewer feedback was positive for the overall flow of the instruments, despite their lengths.
• Interviewers had access to contact details for parent one (based on previous waves), prior to the initial visit to the household. This allowed interviewers to liaise with parent one, as the main point of contact when scheduling an interview. Within the survey, the screener section worked well in reinforcing who parent one in the household was, and whether they were available for the interview.
• Interviewers expressed concerns about some particularly unengaging or uncomfortable questions. They also noted issues with certain questions that were unclear regarding the response options available.
• Interviewers reported some issues using the physical measurement equipment, such as low levels of brightness on the scales or unawareness of how to swap between metric and imperial.

Recommendations for Cohort '08 Wave 7 Main Data collection

The Cohort '08 Wave 7 Pilot data collection successfully tested the distribution of the sample, and efficacy of the data collection instrument in the field. The communication methods with the respondents were evaluated and interviewer feedback was collated on all areas of the data collection process. The following recommendations, based on the pilot outcomes, were suggested to be implemented in the Cohort '08 Wave 7 main data collection.

Sample Distribution Process

• Full sample allocations must be issued to the field force during training, to monitor the syncing of cases between the two tablets.
• Originally all twins and triplets were included in GUI as part of a twin module. For the Wave 7 main collection, only the young person randomly sampled at Wave 1, and their parents, will be interviewed. Completion of the twin module aligns with the transition period of a young person into adulthood. Going forward, the cessation of the twin module will reduce burden on the respondents and time spent in the household.

Interview Length and Content

• The questionnaire should be updated to help reduce its length and reduce participant burden, especially for the YP survey. The suggestions are evidence-based on multiple fronts, including pilot data, interviewer feedback, participant feedback, and para data (Audit Data).
  • Pilot data identified questions with rarely selected response options, which could be condensed into fewer categories.
  • Interviewers reported that the interview felt too long, making them uncomfortable about staying in households for extended periods. They also identified sections where the participants felt unengaged (e.g. education section, food habits) or uneasy to answer (e.g. parents arguing with each other).
  • Participant feedback indicated that they perceived the questionnaire as overly long, which negatively correlated with how favourably they rated it. This suggested that excessive length contributed to participant burden. Participant burden can be negatively related to data quality.
  • Audit data analysis identified several questions that took participants a considerable amount of time to answer, required multiple actions (indicating uncertainty), or had high rates of "Don't Know" or "Refusal" responses.
• For the Principal Survey, other staff members could be invited to answer the survey, if not initially successful in reaching the principal.

Communication with Respondents

• For the main fieldwork, coherent GUI branding across all documentation, website content, and social media must be developed.
• The Cohort '08 data collection and GUI brand will be promoted at the Ploughing Championships, RDS Pregnancy & Baby Fair, and in a poster campaign to schools and Pavee Point.
• The decision was made to add the interviewer’s name and phone number to the initial letter, as well as a link or QR code to the GUI page on the CSO website.