Is there a generic ball ability? An empirical study of measuring and modeling performance on 12 ball tasks.

Conference poster, 2018

Understanding the complexity of motor abilities and skills is a sine qua non for a scientific approach to processes of diagnosis, prediction and program evaluation in human movement and sport performance. However, there is a general paucity of research that seeks to identify the complexity of motor abilities and movement skills in tasks when a ball is involved. Aiming at a contribution towards the formation of a solid evidence basis, this article investigates the interrelationship among performances in various ball tasks. Utilizing a special constructed and technologically enhanced instrument, it is hypothetised that the commonalities of the objectively measured outcomes supports the notion of an underlying generic ball ability. 213 participants were tested in the 12 ball tasks of the instrument (racket juggle, elbow juggle, handwall, footwall, ball machine, knee juggle, board bouncing, bat balancing, roller coaster, foot rolling, tube drop, stick rolling). According to the analysis of the results, 26 of the inter-item correlations were strong (.50-.66), 25 moderate (.31-.49) and 15 modest (.19-.29). Furthermore, latent variable modeling revealed a first order model with one general factor and an excellent fit between model and data (χ² = 62.542, DF = 54; RMSEA = .03, CI90 .00–.05; CFI = .99; SRMR = .03). Ranging between .38 and .80, the loading of the 12 items on the general factor were all significant. Our findings suggest that a general factor, conceptualized as a generic ball ability, underlie participants´ performances on the ball tasks. Due to the popularity of ball games and the usability of ball tasks in various assessment tools, evidence supporting the generality of a ball ability are expected to be further tested as well as inform interventions and valid measurements.