New ethical concerns have been raised about a Scottish Government survey of the health and well-being of school students in 2021-22. This ‘Health and Wellbeing Census’ has already caused controversy, mainly because it asked young people intrusive questions about the life of their families, and also, for people in the fourth year of secondary school and later, about their sexual activity. Many children as young as 15 were thus asked about sex. The survey was carried out in 16 of the 32 local authorities. The others declined to run it because their councillors or officials shared the parental worries. Yet Nicola Sturgeon, in the Times Educational Supplement (9 December 2021), dismissed the objections as potentially ‘whip[ping] up concern on the part of parents for completely unnecessary reasons’.
The new objections go deeper than these troubling questions. The entire ethical basis of the survey is deeply suspect. Drawing on a series of detailed FOI requests made by the chair of the Parental Council of an Edinburgh school, my analysis of the ethics of the survey concludes that the whole exercise is ethically dubious. It concludes also that researchers should not re-use the data because that would compound the ethical offence.
The survey covered the final three years of primary school (P5 to P7) and the full six years of secondary. It was a large project, with responses from 134,044 students, which was 58% of the target population. The questions asked of the respondents dealt with the ways in which ‘attitudes to school, views on neighbourhood and life at home, physical health, and mental health and wellbeing differ by stage, sex and deprivation’. The survey’s very broad interpretation of these aims led to the controversial questions about family life and sexual activity. The students had to answer the questionnaire online in class, supervised by a teacher.
There are six reasons to believe that the survey was unethical.
The first is about consent to take part. The Scottish government cited the ‘public task’ provisions of the General Data Protection Regulation (GDPR) as conferring the authority to undertake the survey on the basis of parental opt-out. That is, consent was assumed provided that it was not withheld. Although this authority is probably legally valid, it is ethically dubious mainly because the survey extended far beyond the routine operation of schools. For the same reason, it may not even be legally valid according to the GDPR.
Numerous reputable research organisations advise that ‘opt-out’ consent is not acceptable ethically. The Scottish Government’s own guidelines on social research that were in place when the survey was undertaken said that ‘an opt-in basis’ should always be the preferred method. The UK Information Commissioner advises that ‘silence, pre-ticked boxes or inactivity’ do not constitute consent. The guidelines of University College London – the UK’s leading university in educational research – advises that ‘failure to opt out does not constitute consent for the purposes of the GDPR’. One reason is that parents were asked whether they wanted to opt out by means of a physical letter sent home with the student. As every parent knows, this kind of letter is easily missed amidst the usual hectic activities of school days.
The second problem is that the ethical and legal right to withdraw consent while responding to the survey was contravened because responses to survey questions that a student had entered before the point of withdrawal remained in the survey data base. This violates basic rights. In the words of the advice from the British Psychological Society, ‘every person from whom data are gathered … should be free during the data gathering phase to withdraw or modify their consent and to ask for the destruction of all or part of the data that they have contributed within agreed and consented limits’.
Third, the arrangement for supervision of school classes while they responded to the questionnaire was unsatisfactory. It did not give adequate specialist support to students who might have difficulty in answering the questions. It also did not give adequate privacy for students who might not want to reveal details to the class teacher or to other students in the class.
Fourth, the process of ethical approval to which the survey was subjected by the Scottish Government did not provide adequate ethical grounds for the choice of individual questions. The dominant ethical rationale for specific questions was that they had been asked in previous surveys done by other organisations. This not a sufficient ethical defence. For example, a question asked in a survey that is independent of government might be ethically acceptable, but not so when the agency doing the enquiry is an arm of government, or is an educational institution that is managerially responsible to government. The questions about sexual activity and family life are prime instances of that.
Fifth, the initial dataset that was compiled from the responses did not protect respondents’ anonymity. The Information Commissioner’s Office ruled in August 2023 that it posed a ‘serious risk’ of harms to students and that ‘the potential harms that may occur are high’. The main reason for that is that the survey identified individual students by their Scottish Candidate Number. These numbers are widely held in schools. They are readily accessible to any teacher or administrator who is involved in arranging the SQA assessments. In some local authorities they are also used as the basis of students’ email addresses (from primary school onwards, long before students are at the stage of taking SQA assessments). That is why the Information Commissioner was alarmed by the implications. Dozens – perhaps hundreds – of people who had access to the Candidate Numbers would, if they also had access to the data from the survey, know private details of identifiable students.
The final ethical problem is about data sharing. Respondents to the survey were not offered adequate information about plans to share the data with organisations outside the Scottish government and local authorities. The Government encourages researchers to link the data from the survey to other information on the people who responded to this survey, but respondents were not given the option of withholding or withdrawing consent for linkages. The ethical problem here is summed up by researchers from Oxford University (writing about consent for data sharing more widely): ‘collecting research data from children and adolescents requires explicit parental consent as well as active assent from the children and adolescents, which can then be further complicated if the study collects data beyond age 16 when the adolescent will need to give their consent anew’.
No-one doubts the importance of research on young people’s well-being. The question is whether that research is done responsibly and with respect for respondents’ rights. The flaws in this Health and Wellbeing survey reveal a governing process that took no more than tokenistic account of the views of students and parents, and that operated with a cavalier disregard of ethical risk.
Lindsay Paterson is Professor Emeritus of Education Policy, Edinburgh University