Skip to main content

Table 3 Range of sample sizes required to estimate PSS-10 effects for a confirmatory RCT comparing aerobic exercise, mindfulness, combination training, and a control arm

From: A 16-week aerobic exercise and mindfulness-based intervention on chronic psychosocial stress: a pilot and feasibility study

Population effect Target MoE N required per cell N adjusted N required per cell per trial Feasible
δz
−0.60 0.30 63 89 9 Yes
−0.37 0.18 132 186 19 Yes
δs
0.80 0.40 59 84 9 Yes
0.70 0.35 73 103 11 Yes
0.60 0.30 96 136 14 Yes
0.50 0.25 133 188 19 Yes
0.40 0.20 202 285 29 *Yes
0.30 0.15 352 496 50 No
0.20 0.10 779 1098 110 No
  1. δz = within-group (T1–T0) gain in PSS-10; δs = between-group differences in gain scores between different arms of a future RCT; MoE = margin-of-error (i.e., the 95% confidence interval half-width); N = number of participant; per cell = paired observations (δz) or observations per arm of a future RCT (δs), Nadjusted = N required per cell adjusted for the retention rate observed in the pilot study. Feasibility is judged against 10 repetitions of the trial over a 5-year period with a maximum of N = 20 per treatment arm across each repetition (i.e., total N = 80 per trial). Yes = trial is feasible. *Yes = trial is not feasible using accuracy in parameter estimation criteria; however, the effect is detectable with 80% power. No = trial is infeasible