TABLE 3.

Computation and meaning of the four essential contrasts in experimental rest-break designs.

	Type	Calculation and interpretation
1	Simple block comparison (relative block difference)	A baseline (no-rest) condition (A) serves to estimate the performance decrement over the testing period without rest breaks. A rests-break (B) condition serves to estimate the performance trajectory when rest is provided. Directly contrasting both yields a measure of the relative A–B block difference in performance, which provides a primitive measure of the overall benefit provided by rest, relative to a continuous condition.
2	Global rest-break effect (relative time change)	To obtain a measure of the “relative” change in performance over the testing period, the trajectory of performance (time-on-task gradient) for both A and B is contrasted. This gives an estimate of the relative change in the time-on-task effect in performance. In other words, it informs how the performance decrement is prevented by rest breaks, relative to when no rest break is given.
3	Local rest-break effect (before–after rest)	The local effect of rest on subsequent performance is obtained by contrasting the adjacent sections before and after the rest break (pre–post rest comparison). This gives an average estimate of the local benefit of rest that immediately occurs in the time series closely before and after taking a rest, irrespective of the time trajectory.
4	Differential effectiveness (early vs. late rest breaks)	To test the assumption that the effectiveness of a rest itself increases with testing time (i.e., with time on task), the local (pre–post) rest effect at different positions during the testing period is directly contrasted. A larger relative effect at late positions in empirical data would indicate that the immediate effect of rest increases over the testing period, in other words, that rest is more effective at late relative to early positions.

The type 1 is, in a strict sense, not interpretable (see Steinborn and Huestegge, 2016) but frequently used in the literature, thus presented here for reasons of completeness. Type 2 tells how rest compensates a potential performance decrement (the time-on-task effect). Type 3 concerns the local dynamics of recovery and thus provides a measure of how immediate recovery occurs directly after the break. Type 4 gives an indication of a change in the local effectiveness change of (early vs. late) rest breaks.