Identifying who is on the edge
Social isolation is quiet and therefore easy to miss. It rarely appears in behaviour data and often does not appear in attainment data until much later. It has to be looked for deliberately.
Ask staff directly
"Which children would you struggle to say anything personal about?" and "Which children do you never see with anyone at playtime?" This produces a different and more useful list than any dataset.
Playground observation
Structured, repeated, at different times. Look for who is alone, who hovers at the edge, who moves between groups without landing, and who stays near an adult.
Simple sociometric information
Careful, anonymous, and used only to inform grouping — never shared with children. Handled badly this does harm; handled well it identifies children nobody had flagged.
Pupil voice
Ask who they would sit with, who they would ask for help, who they play with, whether there is an adult they would go to. Use maps of the school day and scaling rather than open questions.
Cross-check with attendance
Patterned absence — particular days, particular afternoons — is often social rather than academic in origin.
Family view
Parents hear about parties, invitations and playtimes. They usually know before school does.
