Inter- and intra-rater reliability of the Athletic Ability Assessment in sub-elite Australian Rules football players.

The aim of this study was to determine the inter- and intra-rater rater reliability of the Athletic Ability Assessment (AAA) in sub-elite Australian Rules football (ARF) players. Eighteen male ARF players completed the AAA movement assessment (overhead squat, double lunge [left and right], single leg Romanian deadlift [left and right], chin-up and push-up), on two occasions separated by one week. During the first movement assessment players were filmed in the frontal and sagittal planes. Ten raters took part in the study (one experienced rater and nine novices) and were assigned in a quasi-random manner, to complete either (a) real-time assessment on two occasions, (b) real-time assessment on one occasion or (c) video-based assessment on two occasions. When assessed in real-time, of the 7 component movements of the AAA, raters registered moderate or greater intra-rater agreement on between 2 and 5 occasions. Intraclass correlation coefficients (ICC) of between 0.50 and 0.61 for the AAA total score indicated poor real-time intra-rater reliability for this variable. When assessed by video-recording, raters registered moderate or greater intra-rater agreement on between 6 and 7 occasions. The ICC for total score ranged between 0.60 and 0.93. Overall poor inter-rater reliability was evident for AAA component movements regardless of whether it was assessed in real-time or from video. Findings suggest the AAA is most reliably employed when assessed via video. It is recommended that if comparison between multiple raters is desired, a stringent training process be applied so that the interpretation of AAA scoring criteria is standardised across raters.