Fig. 8

Correlation between cases in the community and pupils in November and December. In these panels, a positive lag indicates that the correlation is calculated between schools on the current date and community cases that have been reported at an earlier date up to a maximum lag of 16 days. The correlation is calculated for all LTLAs in each region, rather than calculated individually for each LTLA. For varying time lag applied to data from November and December, the LTLAs presented are a Greater London, b Kent, c Devon and d West Midlands. Secondary schools are depicted by dashed red lines and primary schools by solid blue lines. For each line, the shaded regions indicate the 95% confidence intervals, which were calculated using the Fisher transformation