From arbitrary metrics to meaningful change: Empirical smallest effect sizes of interest for the Feeling Scale during...

Research addressing changes in affective valence during exercise is often difficult to interpret because such changes are typically expressed on arbitrary metrics, such as the widely used Feeling Scale (FS). To determine which changes in FS ratings are perceptible and meaningful and to provide empirically grounded smallest effect sizes of interest (SESOIs), we estimated the minimal perceived change (MPC) and minimal important change (MIC) using global ratings of change (GRoCs) as anchors. Data were collected from 225 participants (M_age = 25.16 ± 8.49, 45% female) who completed two successive cycling sessions at moderate (64-76% HRmax) and light (57-63% HRmax) intensities. The reduction in intensity was designed to elicit more positive affective responses, consistent with dual-mode theory. FS ratings were obtained repeatedly during each session, and perceived changes were assessed retrospectively using GRoC items. Depending on the estimation method, MPC values ranged from 0.26 to 0.77 FS units and MIC values from 0.37 to 0.93 FS units. Given the skewed distribution of GRoC responses, we recommend using the more liberal estimates derived from the mean-change method: 0.77 FS units [95% CI: 0.58, 0.96] for the MPC and 0.93 FS units [95% CI: 0.71, 1.15] for the MIC. These values provide pragmatic SESOIs that can inform a-priori power analyses, support the interpretation of null effects, and allow previously published intervention results to be re-evaluated with respect to their practical relevance. Further research is necessary to validate these estimates, increase their precision, and examine their context dependance.