An 18‐month follow‐up of the Covid‐19 psychology research consortium study panel: Survey design and fieldwork procedures for Wave 6

Abstract Objectives Established in March 2020, the C19PRC Study monitors the psychological and socio‐economic impact of the pandemic in the UK and other countries. This paper describes the protocol for Wave 6 (August–September 2021). Methods The survey assessed: COVID‐19 related experiences; experiences of common mental health disorders; psychological characteristics; and social and political attitudes. Adult participants from any previous wave (N = 3170) were re‐invited, and sample replenishment procedures helped manage attrition. Weights were calculated using a survey raking algorithm to ensure the on‐going original panel (from baseline) was nationally representative in terms of gender, age, and household income, amongst other factors. Results 1643 adults were re‐interviewed at Wave 6 (51.8% retention rate). Non‐participation was higher younger adults, those born outside UK, and adults living in cities. Of the adults recruited at baseline, 54.3% (N = 1100) participated in Wave 6. New respondent (N = 415) entered the panel at this wave, resulting in cross‐sectional sample for Wave 6 of 2058 adults. The raking procedure re‐balanced the longitudinal panel to within 1.3% of population estimates for selected socio‐demographic characteristics. Conclusions This paper outlines the growing strength of the publicly available C19PRC Study data for COVID‐19‐related interdisciplinary research.


| INTRODUCTION
This is the sixth methodological report from the COVID-19 Psychological Research Consortium (C19PRC) Study, a longitudinal, online survey of the UK adult population during the COVID-19 pandemic which was funded by the UK Economic and Social Research Council.
This report aims to provide important methodological information for secondary users of the C19PRC Study data, which is freely available via the Open Science Framework (OSF) (COVID-19 Psychological Research Consortium, 2022). In the next section, we document the context for the sixth wave of the C19PRC Study during August-September 2021, before describing the design and conduct of the survey at this point in the pandemic.

| Context for sixth wave of the C19PRC study
By summer 2021, the UK's hugely successful COVID-19 vaccination programme had built a considerable 'wall of defence' against the coronavirus; approximately 49% of UK citizens aged 18 years and older were vaccinated with two doses of a COVID-19 vaccine, and an additional 17% were partially vaccinated with one dose (Our World in Data, 2021). At that time, emerging evidence from Public Health England (2021) indicated that two doses of a vaccine provided superior protection compared to a single dose against the risk of symptomatic disease and hospitalisation as a result of infection with the now dominant, and highly transmissible, Delta variant. Thus, the UK Government further accelerated the vaccination programme during this period by reducing the interval between vaccine doses from 12 to 8 weeks for those under 40 years, to ensure that all adults aged 18 years and older who wanted to be fully vaccinated had the opportunity to do so by September 2021 (UK Government, 2021b).
Young people aged 16-17 years were also being encouraged to receive their first vaccine dose before the start of the new academic year (Morton & Faulkner, 2021).
As was common throughout the pandemic, each UK nation imposed different regulations to manage the public health crisis (Institute for Government, 2022). In England during July 2021, the Government was working towards step 4 of 'the roadmap'-this document, first published in February 2021, detailed step-by-step plans for the conditions that must be met to permit the relaxation of existing COVID-19 regulations and public health guidance (UK Government, 2021c). Despite on-going concerns about increasing rates of COVID-19 infections and hospitalisations (Ball, 2021;Gurdasani et al., 2021) (see Figure 1), as well as 'pandemic' chaos caused by National Health Service (NHS) app notifications requiring hundreds of thousands of citizens to self-isolate by law due to close contact with a positive case (James, 2021), July 19, 2021 was herald 'Freedom Day' in England. On this day, almost all of the public health guidance and restrictions that had been in place for approximately 15 months were lifted, including: the removal of social distancing measures and limits on social gatherings; mandatory use of face masks in shops and on public transport; work from home orders; and restrictions on operations of nightclubs, theatres, and restaurants (including the requirement for table service in hospitality). Somewhat contradictorily, the UK Government continued to promote a message of caution (UK Government, 2021e), urging: (i) citizens to exercise 'personal judgment' in their interactions with others and to meet up outside where possible; (ii) hospitality to continue with table service where possible; and (iii) employers to facilitate a gradual return to F I G U R E 1 Graphical presentation of the number of daily COVID-19 cases and deaths in the UK, sourced from Our World in Data (2021), aligned to the C19PRC Study survey waves. New daily deaths and cases depicted as 7 day rolling average the workplace (UK Government, 2021a). Many shops continued to require the public to wear face masks and Transport for London ordered that face masks were a 'condition of carriage' (Menendez, 2021). Elsewhere, the devolved governments were more cautious, refusing to reopen fully at this time (Ball, 2021).
On July 1, 2021, it was announced that the UK Government's 'Coronavirus Job Retention Scheme' (CJRS) was set to end on September 30, 2021(UK Government, 2021d. The CJRS was established in March 2020 to provide wages for employees who could not work due to the pandemic or to give financial support to employers to subside the wages of employees in work during the pandemic. Approximately 1.6 million people were covered by the CJRS (or 'furloughed') in July 2021, a substantial decrease from the 9 million working-aged adults who were availing of the scheme in May 2020 (HM Revenue and Customs, 2021). However, the easing of restrictions resulted in a slower-than-anticipated growth of 1.3% in the UK's gross domestic product during quarter 3 of 2021 (July-September), but this was 2.1% below pre-pandemic levels (i.e., of quarter 4, October-December 2019) (Office for National Statistics, 2021). As a result, over one million workers were expected to still require the CJRS when it was due to end in September 2021 (Tomlinson, 2021 Our Consortium's ability to retain a sizeable proportion of the study's baseline sample over multiple waves, coupled with sample replenishment procedures at post-baseline waves and targeted oversampling in the devolved UK nations, ensures that the C19PRC Study is well-positioned to be an authoritative data source from which a public mental health evidence base on the pandemic can be produced . The aim of this methodological report is two-fold. First, we describe in detail the content of, and fieldwork for, this sixth survey wave (hereafter C19PRC-UKW6) to stimulate its secondary use by interested researchers via OSF (COVID-19 Psychological Research Consortium, 2022). Second, we report on: (i) attrition patterns in the C19PRC Study by the sixth wave, and whether these are associated with respondents' socio-demographic characteristics at point of first entry to the study; (ii) sample replenishment and weighting procedures conducted to manage attrition in the longitudinal panel; and (iii) the socio-demographic characteristics and prevalence of probable current common mental health problems (depression,anxiety, related PTSD) among participants in the C19PRC-UKW6 sample.  During Phase 2 (8 to September 28, 2021), new participants were recruited from Qualtrics existing panels to fill quota gaps following the C19PRC-UKW6 Phase 1 recruitment to ensure that the cross-sectional sample for C19PRC-UKW6 was nationally representative of the UK adult population with respect to age, gender, and household income.

| Informed consent process
At all survey waves, participants are informed that their data would be treated in confidence, that geolocating would be used to determine the area in which they lived (in conjunction with their residential postcode stem), and of their right to terminate participation at any time. Participants are also informed that some topics might be sensitive or distressing (e.g., self-harm/suicide content). Information about how their data would be stored and analysed by the research team was also provided. Participants are also informed that they would be re-contacted at a later date to invite them to participate in subsequent survey waves. Participants provided informed electronic consent prior to completing the survey and were directed to contact the NHS website upon completion if they had any concerns about COVID-19, and emotional support services if they had been negatively impacted by any of the questions asked during the survey.

| Compliance with general data protection regulation
At all survey waves, participants are informed that C19PRC data will be stored confidentially in line with general data protection regulation. When the study data is deposited with the OSF, location data is removed and replaced with relevant socioeconomic summary data (e.g., area-level deprivation and population density data). All other personal data is also removed.

| Quality control
As per previous waves , Qualtrics were charged with conducting multiple validation checks on the C19PRC survey data to identify and remove participants who (i) completed the survey faster than the pre-set minimum completion time (to ensure responses were trustworthy); (ii) did not provide consent or meet eligibility criteria; or (iii) did not complete the survey in full. At this wave, 38 respondents from the existing panel were issued the survey as a pilot or 'soft launch' prior to the main fieldwork going live ('full launch') to rectify sequencing/coding errors and omissions. The median survey completion time for the wave is calculated to provide the survey team an opportunity to tailor the content to ensure the median survey time does not exceed 30 min. This is important to minimise respondent burden and maximise participation over time, as well as managing survey costs. The median completion time for the C19PRC-UKW6 soft launch conducted during 5-6 August 2021 was 22 min 51 s. Only minor changes to the survey content were made following the soft launch and these participants were retained in the Phase 1 study sample.

| Measures
An overview of the C19PRC-UKW6 survey content is presented in Table 1.

| Ethical approval
The University of Sheffield provided ethical approval for the C19PRC Study (Reference number 033759).

| Study variables
Given the broad focus of the C19PRC Study in understanding the impact of the pandemic on the UK adult general population, a wide range of socio-demographic, economic, and psychological factors were selected to describe the characteristics of the sample participating at this wave, as well as to identify predictors of attrition from point of first entry into the study (i.e., C19PRC-UKW1, C19PRC-  x Experience of being tested for COVID-19 (symptoms, location of testing/diagnosis) x Experience of waiting to be tested for COVID-19 (self) x Knowing someone close (family member/friend) who has tested positive for COVID-19 x Knowing someone close (family member/friend) who has died due to COVID-19 x Social distancing/hygiene behaviour x COVID-19 vaccination status: Self x

| Data analysis plan and weighting procedures
Five sets of analyses are presented. First, re-contact rates at C19PRC-UKW6 were calculated for Phase 1. Second, a binary logistic

| Outcome of recruitment at Phase 1, C19PRC-UKW6
Following an assessment of the characteristics of non-respondents at

| Weighting procedure for longitudinal panel from baseline (C19PRC-UKW1)
As presented in Supplementary Table S1, the raking procedure successfully re-balanced the characteristics of responders at this sixth wave (N = 1100) to the baseline proportions for gender (rebalance within 1%), age (exact rebalance), household income (within 0.6%), household composition (exact rebalance), urbanicity (exact rebalance), ethnicity (within 0.9%), and born or raised in the UK (within 1.3%).  Applying this weight for all analyses of the C19PRC-UKW6 survey data completed by this longitudinal panel (who entered the panel at baseline) is recommended to account for attrition over survey waves on core study outcomes. Table 4 presents cross-tabulations for the socio-demographic characteristics and the prevalence of probable current common mental health conditions of all adults who participated in C19PRC-UKW6 (N = 2058), stratified by point of entry in the C19PRC Study. Statistically significant differences in characteristics between the C19PRC-UKW6 respondents entering the C19PRC Study at different waves are indicated by different subscripts (see Table 4).

| Socio-demographic and mental health characteristics of adults participating in C19PRC-UKW6
Specifically, adults participating in this wave having entered the study at different time points varied with respect to age, household income, economic activity, ethnicity, presence of children living in household, and experiences of probable current mental health conditions.
For example, participants entering the C19PRC Study at postbaseline waves were typically younger in age (e.g., 5.7% of those who participated in the sixth wave having entered at baseline were 18-24 years vs. 31.8% of those who entered the study at the sixth wave). This was to be expected due to sample replenishment procedures conducted to re-balance the sample following attrition at this wave.
Fewer adults entering the C19PRC Study at post-baseline waves were in the two highest household income brackets, compared to those who entered at baseline (e.g., 14.5% of adults first entering the survey at the sixth wave were earning £57,931+ per annum, compared to 21.0% of those entering at baseline). Higher proportions of adults entering the study post-baseline and participating at this sixth wave were economically activity (e.g., 73.7% at C19PRC-UKW6 compared to 61.5% at baseline), which likely reflects the improving economic environment in the UK following the easing of public health restrictions as the pandemic progressed, as well as the 'top-up' procedures to target younger adults. Higher proportions of adults interviewed at this sixth wave who first entered the C19PRC study at the third or sixth waves were non-white (11.9% and 14.2%, respectively) compared to baseline respondents who re-interviewed at this sixth wave (6.6%). New entrants to the C19PRC Study at this sixth wave lived in households with children (45.5%) compared to those participating in this sixth wave who entered at baseline or the third or fourth waves (25.3%, 32.1% or 23.0%, respectively).
With respect to prevalence of caseness for probable current  -13 of 17 respectively) compared to adults who participated in this sixth wave having first entered the C19PRC Study at baseline or previous postbaseline waves (i.e., probable current prevalence of anxiety, depression, and COVID-19 related PTSD ≤ 22.1% for these respondent groups-see Table 4). All other variables use information collected at the Wave the participant first entered the survey-except ethnicity. Ethnicity was not collected at W3, so information used from W6 only here. *p < 0.05, **p < 0.01, ***p < 0.001. C19PRC-UKW1: baseline survey (March 2020).

| DISCUSSION
14 of 17attrition across a three-wave health-related quality of life during COVID-19 survey in the US between March-April 2020, July-September 2020, and January-March 2021; however, poorer mental health predicted attrition between the first two waves only, but not for the third wave conducted further into the pandemic.
These findings are broadly consistent with our study's current and previous attrition analyses ( . The COVID-19 pandemic has forced a sizeable shift in how research is conducted. Governments across the world have relied on researchers to collect high-quality, high frequency data from members of the general public to help assess and monitor changes in overall health, wellbeing, and socio-economic circumstances to help guide and inform appropriate policy-level responses to support ordinary citizens as they navigated life through the pandemic. The draconian restrictions on social interactions also necessitated the shift towards online survey-based studies, due to the ability of this study design to collect data with greater ease and faster speed when compared to more traditional efforts (Haas et al., 2021;Singh & Sagar, 2021). The long-term impact of the pandemic on the future of survey methodology has yet to be fully understood (Becker et al., 2022;Sattler et al., 2022) and may require novel approaches going forward (Yu et al., 2022).
Recent novel methodological work (Biddle & Sollis, 2021), however, has suggested that survey methodologists should consider routinely asking respondents about their subjective experiences of participating in COVID-19 related (or similar) surveys. For example, in the Australian National University's five-wave COVID-19 Impact Monitoring Survey Program, answers to two questions (i.e., the respondent's subject experience as to how distressing the survey was, and how glad they were that they participated) were strong predictors of attrition over time-that is, those experiencing lower levels of distress while completing the survey and higher levels of gladness having completed the survey were more likely to complete subsequent waves (Biddle & Sollis, 2021). Elsewhere, Yu et al. (2022) reported that self-report difficulty with the COVID-19 health-related quality of life survey predicted attrition across both follow-up waves spanning the first year of the pandemic. Biddle and Sollis (2021) propose that this type of information about participants' subjective experience of survey participation is valuable for tailoring study invitation communi- To conclude, our Consortium has carefully considered and debated the potential impact of the pandemic on our efforts to collect high-quality, longitudinal data from a large sample of the UK adult general population (as well as 'sister' studies in other European countries) using non-probability based sampling methods in detailed methodological reports from each wave of the C19PRC Study . As per best practice (Besançon et al., 2021), we have been strongly committed to Open Science principles from the outset. We encourage interested readers to consult our detailed methodological reports and to access the data via the OSF for exploitation and secondary use purposes when the broad and deep coverage of the C19PRC Study survey data may be suitable for addressing important research questions of public health interest.