Reliable measurement in sport psychology: The case of performance outcome measures

Objectives:
The present research addresses a neglected aspect within the current Zeitgeist of improving methodological standards in (sport)psychology: reliable measurement. We discuss and highlight the importance of reliable measurement from different perspectives and empirically assess reliability of three commonly used performance outcome measures in order to give guidelines to researchers on how to increase reliability of measurements of performance outcomes.
Method:
In three studies we estimate 5 different reliability coefficients for three performance outcome measures based on 14 golf putts (study 1; N = 100), 14 dart throws (study 2; N = 200; 100 sports students; 100 non-sports students) and 14 free throws in basketball (study 3; N = 192; 100 non-basketball players; 92 basketball players).
Results:
The highest reliability was the odd-even reliability for darts for the whole sample (0.888), followed by golf putts (0.714 for distance from the hole, 0.614 for successful putts) and free throws (0.504 non-basketball players; 0.62 for basketball players; and 0.826 for whole sample).
Conclusions:
Based on theoretical considerations and our empirical findings we give practical guidelines to improve reliability for performance outcome measures in sport psychology.