Y t = β 0 + β 1Tt + β 2Xt + β 3XtTt + β 4Zt + β 5ZtTt + β 6At + β 7AtTt + ε t |
T = Time (week) |
X = 0 prior to first lockdown, X = 1 after introduction of first lockdown |
Z = 0 prior to second lockdown, Z = 1 after introduction of second lockdown |
A = 0 prior to end of second lockdown, A = 1 after end of second lockdown |
Β0 = Intercept (predicted weekly count at Week 1: 1–7 January 2019) |
β1 = Period 1 trend (estimated weekly change in outcome during the pre‐lockdown period) |
β2 = Level change at start of first lockdown |
β3 = Difference between period 2 and period 1 trends |
β4 = Level change at start of second lockdown |
β5 = Difference between period 3 and period 2 trends |
β6 = Level change at end of second lockdown |
β7 = Difference between period 4 and period 3 trends |