Table 1.
Longitudinal cohort data.
Fig 1.
Individual growth curves for 25 students per cohort sampled randomly from CITO data.
The grey area represents the lockdown periods.
Table 2.
Results of different ML models with an ITS design for cohort A.
Table 3.
Results of different ML models with an ITS design for cohort B.
Table 4.
Parameters used in the simulation study.
Fig 2.
The dashed line indicates the first intervention and the dotted line the second intervention.
Table 5.
Different models and simulation scenarios.
Fig 3.
Power of the step for simulation scenario with a step and slope change.
Fig 4.
Power of the slope for simulation scenario with a step and slope change.
Fig 5.
Absolute bias of the step for simulation scenario with a step and slope change and an additional step change.
Fig 6.
Absolute bias of the slope for simulation scenario with a step and slope change and an additional step change.
Fig 7.
Power of the step for simulation scenario with a step change, a slope change and an additional step change with correction.
Fig 8.
Power of the slope for simulation scenario with a step change, a slope change and an additional step change with correction.
Fig 9.
Power of the additional step for simulation scenario with a step change, a slope change and an additional step change with correction.
Fig 10.
Power of the step for simulation scenario with a step change, a slope change and a correction for changing slope.
Fig 11.
Power of the slope for simulation scenario with a step change, a slope change and a correction for a changing slope.