When labels and instructions diverge: Operational reclassification of attentional focus in jump performance

Attentional-focus research commonly treats external focus (EF) as a single strategy, although instructions assigned this label may differ in spatial referents and informational content. This study examined whether reliance on author-reported labels was associated with differences in evidence inclusion and meta-analytic inference for jump performance. Five databases were searched, and four prespecified random-effects network meta-analysis (NMA) models were fitted. Model 1 operationally reclassified all 49 eligible datasets at the instruction level. Model 2 retained 41 datasets reported using the generic EF label, Model 3 retained 7 datasets explicitly labelled as proximal external focus (PEF) or distal external focus (DEF), and Model 4 retained 48 datasets while preserving all explicit author-reported categories. Operational review reassigned 42 instructions. In Model 1, PEF, DEF, and holistic focus (HF) improved jump performance relative to control, with standardized mean differences (SMDs) of 0.189, 0.264, and 0.331, respectively. Generic EF estimates in Models 2 and 4 were similar (SMDs = 0.209 and 0.216) and fell between the PEF and DEF estimates from Model 1. Model 3 produced larger but less precise PEF and DEF estimates. Most 95% prediction intervals included the null. Reliance solely on author-reported attentional-focus labels may be associated with differences in evidence inclusion, network structure, heterogeneity, consistency, precision, and resulting conclusions. Future evidence syntheses and systematic reviews should therefore apply prespecified and reproducible instruction-level classification criteria. Primary studies should also report attentional focus more precisely, including the complete wording of each instruction, its intended referent, and relevant information regarding the distance or spatial orientation of that referent.