Monitoring the psychological, social, and economic impact of the COVID‐19 pandemic in the population: Context, design and conduct of the longitudinal COVID‐19 psychological research consortium (C19PRC) study

ABSTRACT Objectives The C19PRC study aims to assess the impact of the COVID‐19 pandemic in the adult population of the UK, Republic of Ireland, and Spain. This paper describes the conduct of the first two waves of the UK survey (the “parent” strand of the Consortium) during March–April 2020. Methods A longitudinal, internet panel survey was designed to assess: (1) COVID‐19 related knowledge, attitudes, and behaviors; (2) the occurrence of common mental health disorders as well as the role of (3) psychological factors and (4) social and political attitudes, in influencing the public's response to the pandemic. Quota sampling (age, sex, and household income) was used to recruit a nationally representative sample of adults. Results Two thousand and twenty five adults were recruited at baseline, and 1406 were followed‐up one‐month later (69.4% retention rate). The baseline sample was representative of the UK population in relation to economic activity, ethnicity, and household composition. Attrition was predicted by key socio‐demographic characteristics, and an inverse probability weighting procedure was employed to ensure the follow‐up sample was representative of the baseline sample. Conclusion The C19PRC study data has strong generalizability to facilitate and stimulate interdisciplinary research on important public health questions relating to the COVID‐19 pandemic.


| INTRODUCTION
Without an effective therapeutic or vaccine aided prophylaxis for the SARS CoV-2 virus (hereafter,, governments around the world have imposed stringent social distancing measures to slow contagion, protect the most vulnerable, and better manage health care service demand/provision (World Health Organisation, 2020a, 2020b. The implementation of such "lockdown" strategies, however, has in turn, unsurprisingly, impacted upon and compromised many facets of social and economic life, resulted in mass unemployment, and sparked fears of an impending global economic crisis and recession (Anderson, Heesterbeek, Klinkenberg, & Hollingsworth, 2020;Hale, Petherick, Phillips, & Webster, 2020). Although the extant literature on the non-biological sequalae of outbreaks of other infectious respiratory diseases (IRDs), specifically SARS-CoV (SARS), the H1N1 flu pandemic, and the Middle East respiratory syndrome (MERS), provides a useful "roadmap" for COVID-19 related research, the unique magnitude of this pandemic, that varies considerably from one country to another (and even across regions within a country), means that the psychological, social, and economic impact of the pandemic is largely unknown.
In March 2020, the COVID-19 Psychological research consortium (C19PRC) was formed to address an urgent need to conduct timely, high-quality research to generate a robust evidence base that could be used by policymakers and clinicians to successfully navigate the rapidly evolving COVID-19 crisis. The C19PRC comprises psychologists, mental health researchers, and political scientists, and involvement from non-academic stakeholders in the disciplines of public health and social care, from across the UK, the Republic of Ireland, and Spain. The C19PRC's chief aim is to monitor and assess the long-term psychological, social, and economic impact of the pandemic, and to investigate how this might vary across countries that differed in relation to their public health response strategies. Two core principles directed that research conducted by the C19PRC would be: (1) informed by previous studies that investigated the non-biological consequences of previous outbreaks of IRDs and on recommendations from researchers and public health representatives in response to these outbreaks; and (2) responsive to the unique socio-political-economic landscape of the countries involved, and to prioritize the exploration of between-country factors that might help explain differences in COVID-19 related outcomes. What follows is an overview of the preparatory work conducted to devise the C19PRC's research methods framework.
Unsurprisingly, the National Advisory Committee on SARS and Public Health (Naylor et al., 2003) proposed that a "systemic perspective" was needed and should be prioritized by those engaging in IRD psychosocial research. This recommendation suggested that psychosocial research should not be restricted to health care workers and patients during such crizes and that populations such as nonmedical personnel and the public should also be assessed. Such an approach would enable more comprehensive and balanced planning of efforts to alleviate the psychosocial burden of IRDs in the population at large or mitigate its onset in the future (Sim & Chua, 2004 The extant evidence base on outbreaks of IRDs also provided valuable direction in relation to a variety of other relevant issues that warrant investigation during the pandemic. Significant attention has also been paid to the role, context and change in public health knowledge, attitudes, behaviors, and practices over the course of IRD outbreaks (Alsahafi & Cheng, 2016;Cowling et al., 2010;Karademas, Bati, Karkania, Georgiou, & Sofokleous, 2013;J. Lau, Griffiths, Au, & Choi, 2011;J. T. Lau, Griffiths, Choi, & Tsui, 2010;Lin et al., 2011).
1.4 | Principle 2. C19PRC: Prioritizing a multicountry approach to researching the impact of the pandemic.
Between-country differences in COVID-19 related outcomes, although clearly evident in the earliest surveillance studies (World Health Organisation, 2020a, 2020b), are not easily explained, and may be related to a host of macro and micro-level factors including: (1) the size and geographical location of a country; (2) the characteristics of a country's population (e.g. age/sex/health conditions distribution); (3) the level of population density within a country; (4) the ability of a country's health service to conduct widespread testing for COVID-19, to initiate and maintain a robust contacttracing system for confirmed COVID-19 cases, and to provide optimal care for individuals diagnosed with COVID-19; (5) the timing, nature and severity of "lockdown" restrictions (and the subsequent implementation of policies and procedures to "unwind" such restrictions); and (6) the public's motivation and ability to comply with the lockdown restrictions, as well as their attendance to, and engagement with, national public health initiatives enacted to prevent the spread of the virus (Hale et al., 2020, p. 31;V. J. Lee, Chiew, & Khong, 2020;Olagnier & Mogensen, 2020;Wyper et al., 2020).

| C19PRC objectives
Armed by the research evidence underpinning the C19PRC's two core principles, five objectives were set for the Consortium's research methods framework: (1) to recruit a large, nationally representative panel of adults in each country, that will serve as a study "spine" in that country, and that will facilitate various opportunities for conducting bespoke studies targeting specific sub-groups in the population (e.g. frontline workers) who are particularly at risk for COVID-19; (2) to collect detailed "baseline" survey data on a broad range of outcomes and behaviors, such as mental health disorders (e.g., anxiety, depression) and health-related behaviors (e.g., maintaining hygiene practices, face-mask wearing), known to be impacted by, and influence recovery from, a global pandemic; (3) to assess a broad array of protective and risk factors known to (or thought to) influence identified health-related outcomes and behaviors, both at the micro-level (i.e., via respondent self-report) and macro-level (i.e., via linkage of respondent geospatial data to external data resources e.g. population density; availability of green spaces; area-level rates of COVID-19 infection, etc.); (4) to re-contact respondents regularly as the pandemic unfolds, with measurements at each assessment being guided by both the extant literature on previous pandemics, but also adapting and responding to, the unfolding social, political, and economic circumstances in each country; and (5) to produce rapid, high-quality, country-specific research outputs in the first instance, for the purpose of contributing to the emerging evidence base pertaining to the pandemic at a national level, and also, to prioritize multi-country research outputs to highlight, more broadly, whether and how between-country differences might uniquely explain variation in COVID-19 outcomes.
The C19PRC prioritized consistency in the measurement of core study outcomes (e.g. mental health disorders; engagement in COVID-19 health-related behaviors) across study countries, collection of baseline data at the earliest opportunity in each country, and also timely, roughly equal follow-up data collection in each country.
Flexibility to tailor sections of the survey to address and assess country specific issues (e.g. differences in "lockdown" strategies) was also recognized as an important feature of the study.
The C19PRC study was designed as an internet-based nationally representative panel survey. The first C19PRC survey commenced in the UK in mid-March 2020, 52 days after the first case of  had been confirmed in the UK. The first wave of this UK survey (hereafter, C19PRC-UKW1) is the "parent" survey of the C19PRC Study; subsequent rollouts in other countries (so far, the Republic of Ireland and Spain-both of which commenced in late-March/early April 2020), were modeled on the design of the C19PRC-UKW1. The remainder of this paper describes the content and conduct of the C19PRC-UKW1 and the follow-up survey at wave 2 (hereafter, MCBRIDE ET AL. -3 of 16 C19PRC-UKW2). While the paper serves as a broad methodological overview for the first and follow-up surveys in all three countries, minor differences in methodologies across countries will be described in country-specific papers (e.g. see Hyland et al. [2020] [Valiente, 2020]). Forthcoming methodological papers will detail our international Consortium's progress throughout 2020-21, including the conduct of baseline surveys in other countries (e.g. Italy) and follow-up waves in the UK, Republic of Ireland and Spain, as well as studies of a qualitative and experimental design linked to the UK parent strand. Qualtrics. Qualtrics has completed more than 15,000 projects across 2,500 universities worldwide.

| C19PRC-UKW1 sampling design and procedure
The UK adult population aged 18+ years was the target population for C19PRC-UKW1. Quota sampling methods were employed to achieve a representative sample in terms of age and sex (using 2016 population estimates from Eurostat [2020]) and household income (using 2017 income bands from the Office for National Statistics As an aggregator of panels, Qualtrics provides the online platform to securely house data and leverages partners to connect with respondents. Qualtrics recruits study participants from traditional, actively managed, double-opt-in market research panels, that are used for corporate and academic market research only. Potential respondents were alerted to the C19PRC-UKW1 by Qualtrics in one of two ways: (1) they opted to enter studies they were eligible for by signing up to a panel platform; or (2) they received automatic notification through a partner router that alerted/directed them to studies for that they were eligible (either via email, SMS, and in-app notifications). Importantly, to avoid selfselection bias, survey invitations to eligible participants only provide general information and did not include specific details about the contents of the survey. Participants were required to be an adult, able to read and write in English, and a resident of the UK. No other exclusion criteria were applied. Panel members were not obliged to take part; however, members routinely receive an incentive for survey participation based on the length of the survey, their specific panel list profile, and target acquisition difficulty, among other factors. The specific type of reward varies and may include cash, air miles, gift cards, redeemable points, charitable donations, sweepstakes entrance, or vouchers.
Qualtrics proceeded as follows during the six days of fieldwork to fill the quotas (1) respondents in "hard to reach" quota groups (e.g. young adults in the highest income bands) were targeted first; (2) next, the focus shifted to allow the quotas to "fill up" naturally; and (3) finally, a switch back to targeting respondents to fill incomplete quotas ensued. Those who chose to participate followed a link to a secure website and completed all surveys online. The invite link was active for a participant until a quota they would have qualified for was reached but after this time, previously eligible respondents were prevented from taking part.
A power analysis for the C19PRC-UKW1 sample size was con-

| Informed consent process
Participants were informed about the purpose of C19PRC Study, 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 without giving a reason. Participants were also informed that some topics may be sensitive or distressing. Information about how their data would be stored and analyzed by the research team was also provided. All participants provided informed consent prior to completing the survey and were directed to contact the NHS 111 helpline upon completion if they had any concerns about COVID-19.

| Compliance with General data protection regulation (GDPR)
C19PRC data will be kept confidential in line with GDPR. In accordance with GDPR, contact details were separated from the dataset and personal data is restricted to members of the research team.
When the study data is to be deposited with the UK Data Service (see Discussion), location data will be removed and replaced with relevant socioeconomic summary data (e.g. area-level deprivation and population density data). All other personal data will also be removed.

| Quality control
A minimum survey completion time was set at 11 min, 11 s (i.e., half the based on soft-launch median completion time of 22 min, 22 s).
Qualtrics employed checks to identify and remove any participants who (1) completed the survey in less than the minimum completion time (to ensure responses were trustworthy) or (2) were potentially duplicate respondents. Median completion times were 28.91 min for C19PRC-UKW1 and 36.15 min C19PRC-UKW2.  Table 1 for an example as to how this was conducted in the UK).

| Ethical approval
Ethical approval for the project was provided by the University of Sheffield (Reference number 033759).

| Data analysis plan and weighting procedure
Four sets of analyses are presented to (1) demonstrate the success of the quota sampling methodology at C19PRC-UKW1; (2) determine the representativeness of the C19PRC-UKW1 sample to the UK population for a suite of socio-demographic characteristics not used for quota sampling; (3) describe the sociodemographic characteristics of the C19PRC-UKW2 sample, as well as the characteristics of those lost to follow-up; and (4) estimate the extent to that an a priori selection of socio-demographic characteristics (age, gender, income, urbanicity, household composition, and having been born/raised in UK) predicted attrition at C19PRC-UKW2 using a binary logistic regression analysis.

| Quality control checks and representativeness of C19PRC-UKW1 sample as per quota sampling methods
Given the dual processes used by Qualtrics and partners to recruit respondents to quotas, it was not possible determine the number of survey invites that were distributed to panel members, or indeed the number of panel lists who were alerted to the survey and who did/did not complete the survey (i.e. a response rate).
Qualtrics provided metrics for the C19PRC-UKW1, as follows: (1) having original commenced the survey, 159 respondents did not provide full informed consent and were screened out; 35 respondents who completed the survey from outside the UK or were aged under 18 years (n = 6) were also screened out; to ensure responses were trustworthy, 77 participants who completed the survey in less than the minimum completion time were removed, as were 64 potential duplicate respondents. This resulted in a C19PRC-UKW1 sample of 2025. Table 2 compares the prerecruitment quotas to those achieved during the fieldwork period.
Sex quotas were obtained to within 1% (slightly more women than men were recruited), age quotas were obtained to within 0.1%-0.6% (fewer respondents aged 25-44 years were recruited), and household income band quotas were obtained to within 0.25%-1% (fewer respondents in the middle income band £25,341-£38,740 were recruited).

| Representativeness of the C19PRC-UKW1 sample-UK population estimates.
Reliable estimates for some population estimates (e.g. age, and sex) can be obtained annually from non-census sources (Office for National Statistics, 2019a). Despite on-going efforts to develop methods to produce similar reliable mid-census population estimates for characteristics such as ethnicity (Office for National Statistics, 2019b), the 2011 census remains the most reliable source of population data for many socio-demographic characteristics (e.g. ethnicity, and country of birth) (Office for National Statistics, 2020), despite MCBRIDE ET AL. -7 of 16 the recognition that the population structure has changed since 2011. Ireland were not feasible-see Table 3).

| Follow-up of C19PRC-UKW1 sample at C19PRC-UKW2
A total of 1508 respondents initiated the C19PRC-UKW2 interview (overall re-contact rate of 74.5%) but 102 respondents did not complete the survey in full. Of these, the majority were female  -9 of 16 males, older people, those in higher income brackets, those of White ethnicity, people living outside cities, those living in lone adult-only households, and individuals born/raised in the UK. Subsequently, an inverse probability weighting procedure (Seaman & White, 2013) was conducted to ensure that the C19PRC-UKW2 sample was representative of the C19PRC-UKW1 sample.
T A B L E 3 Comparison of representativeness of the COVID-19 psychological research consortium (C19PRC) study UK Wave 1 (C19PRC-UKW1) sample to UK adult population for key socio-demographic characteristics, by country, March 2020 (N = 1951)  N = 703,275), not for all adults aged 18+ years, and therefore a comparison to survey for household composition is not feasible.

Comparison to UK adult population (+/−% difference between survey sample and population)
T A B L E 4 Sociodemographic characteristics of respondents, COVID-19 psychological research consortium (C19PRC) study UK Wave 1 (C19PRC-UKW1), March 2020 and Wave 2 (C19PRC-UKW2), April 2020 emphasized recently that now, more than ever, research studies must be of high quality (Nieto, Navas, & Vázquez, 2020) and that this can only be achieved through the use of focused questions, the employment of robust methodologies and the securement of necessary ethical approval(s) from relevant institutions (Hipp, Bünning, Munnes, & Sauermann, 2020;Townsend, Nielsen, Allister, & Cassidy, 2020;World Health Organisation, 2020c). We believe that, through the design and initiation of a robust, representative, longitudinal, multicountry study, during the earliest phases of the SARS-CoV-2 pandemic, the C19PRC study surpasses these research quality standards.
Several strengths of the C19PRC ensure that this study's data will be well placed to make a significant contribution to the knowl-  (6) securing permission to re-contact members from the survey "spine" for additional "sub-projects". Such projects, to include methodologies such as qualitative interviewing, experimental designs, and experience sampling methodologies, as well as other "spin-off" studies involving adolescents and young people (University of Sheffield, 2020), will provide a unique opportunity to apply a mixed-methods approach to uncover important aspects of the pandemic, as they unfold over time, in greater detail.
As is common with all studies, the C19PRC study is not without limitations and chief among these is the use of quota sampling to recruit the non-probability based sample via the internet. This optin mode of recruitment employed by Qualtrics, albeit being a costeffective method for gaining fast access to a large and diverse sample (and the only feasible method of recruitment during the pandemic), inevitably meant that it was not possible to generate a response rate for the baseline survey due to the lack of a known denominator or sampling frame. While more research is required to fully investigate the strengths and weaknesses associated with internet-based panel surveying (Bergeson, Gray, Ehrmantraut, Laibson, & Hays, 2013), it has been suggested that the composition of non-probability internet-based survey panels differs from that of the underlying population (Hays, Liu, & Kapteyn, 2015 Opinion Research (APPOR) asserts that when non-probability sampling (as opposed to probability sampling) methods are used, there is a higher burden of responsibility on investigators to describe the methods used to draw the sample and collect the data, so that users can make an informed decision about the usefulness of the resulting survey estimates (Baker et al., 2013). We support the APPOR's position that it is useful to think of different nonprobability sampling approaches as falling on a continuum of expected accuracy of the survey estimates; at one end are uncontrolled convenience samples that produce risky survey estimates by assuming that respondents are a random sample of the population, whereas at the other end, there are surveys that recruit respondents based on criteria related to the survey subject matter and then the survey results are adjusted using variables that are correlated with the key study outcome variables (Baker et al., 2013). The design of C19PRC ensures that it falls towards the latter end of the continuum. Despite the demonstrated representativeness of the C19PRC-UKW1 sample, the Consortium are cognizant that the sample composition recruited at baseline and retained at first follow-up may make specific sub-group analyses (e.g. between-country; potentially vulnerable groups of the population, such as frontline/key workers and/or ethnic minority respondents) difficult. Our plans to recruit additional survey members into the study in subsequent waves of the UK parent strand, with a specific focus on over-sampling respondents in Wales, Scotland, and Northern Ireland, and among underrepresented subgroups of the population, where possible, shall facilitate more robust comparisons of this nature for the core study mental health outcomes. Details of this methodological approach will be described in forthcoming methodological papers from the C19PRC study.
The C19PRC study data will facilitate and stimulate interdisciplinary research on important public health questions such as: (1) What role does the public's knowledge, attitudes, behaviors, and practices have in determining health outcomes during the COVID-19 pandemic? (2) What level of trust does the public have in public/ political institutions and how is this associated with compliance with COVID-19 related health/protective/preventative behaviors or practices? (3) What is the psychological impact and sequalae of COVID-19 and its associated socio-economic effects in the UK? (4) Who is most at risk of psychological distress during COVID-19? (5) What does resilience look like in the context of COVID-19 and what factors contribute to it? and (6) How do the public feel about future vaccination for COVID-19? Findings from the C19PRC study on the topics of mental health disorders and COVID-19 related health behaviors have already been accepted for publication (Gibson Miller et al., 2020;Shevlin et al., 2020). The C19PRC has recently received funding from the UKRI/ESRC COVID-19 rapid response call, supporting on-going data collection for the parent strand of the study into mid-2021. Finally, our consortium is committed to depositing our survey data with the UK Data Archive Service and to actively and widely encouraging re-use of the data so that maximum benefit can be achieved from this robust data resource.