Table 1 lists the elicitation questions relating to contact numbers and table 2 lists questions related to risk mitigation. The full questions and data for completed responses are provided in electronic supplementary material, S1 (§1).

Quantitative elicitation questions related to contacts (see §1 of electronic supplementary material, S1 for full questionnaire). The Table format is as sent to the experts with some words in bold or underlined to give emphasis to help with the clarity if the questions.

aIn this and subsequent similar questions you are being asked about the typical or average behaviour and your uncertainty in this average or typical behaviour.

bThis question and subsequent similar questions you are asked to think about the extreme behaviours of individuals in your school.

c It seems likely to us that there will be little difference in the behaviour and management regime for very young children but this is an opportunity for you to disagree.

dNumbers, character and organization of Year 2 to 5 (emergency workers and vulnerable children) may differ from Year 6 affecting contacts. This question might not be relevant to some schools.

Questions related to risk mitigation.

Because epidemiological models use contact data as a basis for modelling transmission of infection [5,7,19], the majority of our elicitation questions focus on contacts between persons in schools. The greater the number of contacts and the longer the duration of those contacts, the greater the chance of infection transmission [19]. In this study, a contact is defined as a conversation or interaction at a spacing of 1 m or less for 5 min or more. Previous studies of contacts in elementary/primary schools [7,8], using personal position wireless mote devices, found that the relationship between contact numbers and duration exhibits a power-law distribution. Those studies, acknowledging the arbitrariness, had chosen contacts of 5 min or more as significant, to be counted towards daily contact totals. However, mote devices can detect contacts at distances of up to 3 m depending on local transmission paths affecting signal strength between devices. Thus, our definition of a contact is, on average, generally comparable with results from these previous studies. We also justify this choice a posteriori by showing the contribution of short contacts only add modestly to total contact duration, which is dominated by long-duration contacts.

We asked two kinds of questions to characterize daily contacts for individuals within a cohort. One kind of question aimed at estimating daily contact counts for a typical representative individual on a normal pre-COVID day (Q2a, Q5a, Q8a and Q10a) and then in ‘new normal’ times (Q3a, Q6a, Q9a and Q11a). The other kind of question aimed at estimating variations between different individuals in terms of their daily contacts on a pre-COVID day (Q2b, Q5b, Q8b and Q10b) and in new normal times (Q3b, Q6b, Q9b and Q11b). In these latter questions, experts are asked to think about the variation of contacts for different individuals within the cohort. Here ‘least’ and ‘most’ are used to make comparison with the median values elicited in the first kind of the question for the ‘typical’ individual. Combining the results for the 50th percentile value from questions of the first kind with values of ‘least’ and ‘more’ from questions of the second kind provides a measure of the variability of contacts for individuals in a cohort; this represents a quantitative realization with meaning for epidemiological modelling.

By contrast, Questions Q12–Q15 concerned contacts between individuals who are members of different cohorts. These questions were augmented by questions on bubble sizes (Q1b) and on contact rate comparisons between children of different age (Q4 and Q7). Most questions required elicitation of a range to reflect epistemic and aleatory uncertainties, which are discussed in more detail in the Results sections below. Questions were also asked on risk mitigation measures (Q1a, Q16–Q24) which required a mixture of quantitative and qualitative responses.

Before circulating the calibration and elicitation questions to the teachers, the questions were reviewed by Prof. Andrew Noyes (Faculty of Education, University of Nottingham) to check their clarity. Some revisions were made as a consequence.

Normally, experts are convened in a plenary meeting or workshop, in order to conduct the elicitation in a facilitated, structured manner. Such a meeting usually covers: introduction to the methodology; calibration of the experts using seed questions; presentation and discussion of the questions; and a time for the experts to answer the questions. Depending on the scope, complexity and number of issues to be elicited, a meeting might typically last from one to three days, with time allowed for discussion on the evidence that can inform individual experts' responses to the questions. After the experts’ responses have been processed, it is customary to discuss the findings with the participating group; usually, there is an opportunity for participating experts to critique results and, where necessary, for some questions to be clarified and any disputed critical items re-elicited.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.