Introduction
McPherson et al. (2021) examined psychological wellbeing among adults in the United Kingdom during the first national COVID-19 lockdown through a three-wave online longitudinal survey. The study followed 1,958 adults and focused on trajectories of anxiety, depression, and COVID-19-related traumatic stress rather than treating the population as though everyone experienced the pandemic in the same way. Growth mixture modelling was used to identify groups with different symptom patterns, and logistic regression was then used to examine factors associated with membership in those groups. The analysis identified four broad trajectories: low and stable symptoms, high and stable symptoms, symptoms that improved, and symptoms that worsened across the study period. Mental-health characteristics, sociodemographic factors, and COVID-19 worries were among the predictors associated with trajectory membership. Because the researchers measured rather than assigned pandemic experiences, the project is an observational longitudinal study, not an experiment. Its value lies in showing heterogeneity over time while also illustrating the ethical and methodological responsibilities involved in collecting sensitive psychological data remotely during a public-health emergency.
Research Design, Variables, and the Limits of Causal Language
The study’s repeated-wave design is stronger than a single cross-sectional survey because it can distinguish stable distress from improvement or deterioration over time. Anxiety, depressive symptoms, and COVID-19-related traumatic stress function as measured outcomes, while prior mental-health characteristics, loneliness, trauma exposure, demographic variables, health circumstances, and pandemic-related worries can be examined as predictors of different trajectories. Growth mixture modelling and logistic regression are methods of analysis rather than methods of data collection; the data themselves came from repeated online questionnaires. This distinction matters because a statistical model cannot convert an observational design into a randomized experiment. Participants were not assigned to a lockdown, level of worry, or health condition, and many correlated influences were operating simultaneously. Consequently, the results support statements about trajectories and associated risk or protective factors but do not establish that one measured predictor independently caused an individual’s symptoms to rise or fall. The most defensible interpretation is that the three-wave design revealed meaningful patterns of psychological response and identified variables associated with greater or lower vulnerability, which can guide hypotheses and support planning for future emergencies (McPherson et al., 2021).
Sampling, Digital Access, and Representation
Recruitment through social media and the Prolific online participant platform made rapid data collection possible when face-to-face research was restricted, but convenience and online-panel recruitment create important questions about representation. People without reliable internet access, sufficient digital literacy, adequate English proficiency, spare time, or familiarity with online research may have been less likely to participate. Individuals interested in psychology or already accustomed to online studies may have been more willing to enroll, while people experiencing severe illness, acute distress, caregiving pressure, or unstable housing may have been less able to complete repeated waves. Attrition can introduce an additional distortion if participants who remain differ systematically from those who stop responding. These limitations do not make the data invalid or “inauthentic”; online self-report can reduce interviewer pressure and may facilitate disclosure of sensitive experiences. The correct methodological response is to describe the sample carefully, compare retained and lost participants where possible, and avoid treating the participants as a perfect miniature of all UK adults. Ethical research also requires attention to digital exclusion because convenient recruitment methods can systematically leave some populations underrepresented.
Informed Consent and Psychological Risk in Remote Research
Electronic consent can be ethically valid when participants receive clear, understandable information about the study’s purpose, procedures, duration, repeated contacts, foreseeable discomfort, compensation, confidentiality, withdrawal rights, and research contacts. A checked box should document consent rather than substitute for a meaningful consent process. Federal guidance on electronic informed consent emphasizes that researchers remain responsible for providing information and opportunities for questions even when consent occurs remotely (U.S. Department of Health and Human Services & U.S. Food and Drug Administration, 2016). This responsibility is especially important in a survey asking about anxiety, depression, loneliness, trauma, and pandemic-related fear. Such questions can cause discomfort even when a protocol is classified as minimal risk. Participants should know that they may stop participation and, where scientifically appropriate, skip sensitive items without unreasonable penalty. Researchers also need a predetermined plan for mental-health and crisis resources, including the limits of researcher monitoring. If responses are not reviewed in real time, participants should not be led to believe that completing a questionnaire is equivalent to receiving clinical assessment or emergency support.
Privacy, Longitudinal Linkage, and Data Security
Psychological survey data are sensitive, and a longitudinal study creates additional privacy challenges because researchers must connect a participant’s responses across several waves. That linkage means data should not be described as fully anonymous when a code, account, email address, or other mechanism allows records to be reconnected. A more accurate distinction is between identifiable data and confidential research data protected through technical and organizational safeguards. Contact details should be separated from response files wherever feasible, access should be limited to authorized personnel, and data should be encrypted in transit and storage. Researchers should also disclose whether a third-party survey service or recruitment platform receives identifiers or usage information. The consent materials should explain retention periods, future data sharing, and any limits on withdrawal after de-identification or aggregation. The Office for Human Research Protections provides continuing guidance on informed consent and protection of research participants, reinforcing that remote methods do not reduce an investigator’s responsibility for confidentiality (Office for Human Research Protections, 2026). Publications should report aggregate patterns and avoid combinations of characteristics that could re-identify small or distinctive subgroups.
Measurement, Attrition, and Responsible Interpretation
Validated psychological scales improve consistency, but questionnaire scores remain measures of symptoms rather than automatic clinical diagnoses. Responses may be influenced by recall, mood at the time of completion, social desirability, household privacy, interpretation of questions, or changes in circumstances between waves. Longitudinal attrition is equally important because participants with worsening symptoms might withdraw, while those whose lives stabilize may become less motivated to continue; either pattern can change estimated trajectories. Researchers should therefore report missing-data procedures, examine whether attrition is associated with baseline characteristics, and avoid assuming that missingness is random without evidence. Mixture models also require interpretive caution because the number and shape of latent classes depend partly on modelling decisions, fit criteria, and substantive judgment. The four-class result in this study is informative because it rejects a single narrative of universal decline, but trajectory labels should not be treated as permanent personality types. Ethical communication distinguishes temporary symptom patterns from psychiatric diagnoses and reports uncertainty alongside averages, subgroup differences, and model-based classifications.
Scientific Strengths and Practical Significance
The study’s major strength is temporal depth during a rapidly changing period. Three waves allowed the researchers to observe whether distress remained low, stayed elevated, improved, or emerged later, providing a more useful picture than a one-time estimate of average anxiety or depression. The sample of 1,958 adults was substantial, and the use of established measures and trajectory modelling enabled the authors to investigate heterogeneity that a single mean score could conceal. These strengths have practical significance because mental-health support during emergencies should not assume identical needs at identical times. Some people may require sustained support, others may experience early distress that improves, and another group may develop difficulties after an initial period of coping. At the same time, services should not be allocated solely from a predictive model created in one online sample. Findings should be combined with clinical assessment, local service data, social conditions, and evidence from other studies. The study is therefore most useful as evidence about patterns and associated vulnerability rather than a diagnostic instrument or a proof that any one pandemic factor produced a particular psychological outcome.
Conclusion
The UK COVID-19 Psychological Wellbeing Study offers a strong example of how longitudinal online research can reveal diverse mental-health trajectories during a public-health emergency. Its three-wave design, large sample, validated symptom measures, growth mixture modelling, and analysis of risk and protective factors provide richer evidence than a single cross-sectional survey. The study also demonstrates why methodological accuracy and research ethics must remain closely connected. Growth mixture modelling and logistic regression analyze data; they do not collect it, and observational associations should not be described as experimentally proven causes. Remote recruitment increases speed and feasibility but raises questions about digital exclusion, representation, attrition, privacy, and the meaning of electronic consent. Sensitive psychological measures require clear warnings, withdrawal options, confidentiality protections, appropriate resources, and careful communication that symptom scores are not diagnoses. When these safeguards and limitations are made explicit, online longitudinal research can contribute valuable knowledge without overstating certainty. The central finding is not that every UK adult deteriorated psychologically, but that people followed different symptom trajectories and may therefore require different forms and timing of support.
References
McPherson, K. E., McAloney-Kocaman, K., McGlinchey, E., Faeth, P., & Armour, C. (2021). Longitudinal analysis of the UK COVID-19 Psychological Wellbeing Study: Trajectories of anxiety, depression and COVID-19-related stress symptomology. Psychiatry Research, 304, 114138. https://doi.org/10.1016/j.psychres.2021.114138
Office for Human Research Protections. (2026). Guidance and informed consent resources. U.S. Department of Health and Human Services. https://www.hhs.gov/ohrp/regulations-and-policy/guidance/index.html
U.S. Department of Health and Human Services & U.S. Food and Drug Administration. (2016). Use of electronic informed consent: Questions and answers. https://www.hhs.gov/ohrp/regulations-and-policy/guidance/use-electronic-informed-consent-questions-and-answers/
Academic Master Education Team is a group of academic editors and subject specialists responsible for producing structured, research-backed essays across multiple disciplines. Each article is developed following Academic Master’s Editorial Policy and supported by credible academic references. The team ensures clarity, citation accuracy, and adherence to ethical academic writing standards
Content reviewed under Academic Master Editorial Policy.
