Goal-setting in sport: across 36 studies it improves performance by 0.34 SD
Moderate, absolute, and combined short- plus long-term goals work best; "harder is always better" does not carry over from industry
Key Findings
Goal-setting reliably improves performance in industrial psychology, but the effect had not been clearly shown in sport; the authors argue earlier sport studies lacked statistical power and pool 36 studies meeting inclusion criteria.
Overall, setting goals improved sport performance by 0.34 of a standard deviation.
Moderate goals, absolute goals, and short-term goals combined with long-term ones were linked to the largest effects — i.e. the industrial "the harder the goal, the better" gradient (Locke & Latham 2002, DOI 10.1037/0003-066X.57.9.705; not checked here) did not hold cleanly for physical tasks.
Additional moderators were identified (full text; not checked against the abstract): goals set with athlete participation and goals made public showed larger effects than assigned or private goals.
Not from this paper: McEwan 2016 (DOI 10.1080/17437199.2015.1104258; not checked against its abstract) reports that multi-component goal-setting interventions improve physical-activity behaviour.
Practical Application
Across 36 sport and exercise studies, goal-setting raised performance by about a third of a standard deviation, and moderate, absolute and short-plus-long-term goals worked best. Make goals specific and measurable — a concrete power or pace held for a set duration rather than "ride hard" — and pair a controllable process goal with the outcome goal.
Set realistic-but-stretching (moderate) targets; avoid goals that are failures in advance, since very hard goals did not outperform moderate ones here.
Involve the athlete and make goals semi-public (full-text moderators); nest session and weekly goals inside the season goal and give feedback against them (REF-630 for the self-monitoring loop).
Limitations
1995 meta-analysis of older, often lab or short motor tasks; "moderate > hard" is within physical tasks (cognitive goals still follow the Locke–Latham gradient); moderate pooled effect.
How this entry was checked and changed
Content correction 2026-09-27
"Public and athlete-participated goals beat assigned/private" labelled as full-text moderators (the abstract names only moderate, absolute and combined short/long-term goals); McEwan 2016 marked as not checked.
Illustrative "250 W for 20 min" replaced with a number-free example; Sport line normalised to "general".