Abstract
Sport performance is an important and complex multidimensional and context-influenced factor that is often misunderstood or oversimplified. However, sport performance is often haphazardly measured and often assessed through the application of non-sport measures (e.g., Game Performance Assessment Inventory) and/or observable metrics (e.g., time), highlighting a need for a context-specific perception-based measure. The purpose of this research was to develop and evaluate context-relevant measures of sport performance perceptions. An initial set of 34 items was generated reflecting elements of sport performance perceptions, the Long-Term Athlete Development Model, and athlete/coach health metrics. Athletes (N=197, women=93, men=102, 2=non-binary gender, Mage=23.16) from a range of competition levels and sport types, were invited to complete a brief online survey. Athletes (n=88) also completed a follow-up of the items 5 to 25 (M=18.77 weeks) weeks later. A set of two, 12-item measures resulted from a Confirmatory Factor Analysis: 1) The Sport Performance Perceptions Scale-Training and 2) The Sport Performance Perceptions Scale-Competition. Model fit indices for Training were χ²(54)=105.04, CFI=.87, TLI=.84, RMSEA=.072, SRMR = .061; and for Competition were χ²(54)=140.08, CFI=.84, TLI=.80, RMSEA=.093, SRMR=.070.Test-retest results for Training Time (α=0.77) versus Time (α=0.73) were r=0.455; p<.001, while Competition Time (α=0.81) versus Time (α=0.73) were r=0.694; p<.001. These fit indices, in addition to the test-retest reliability, and ecological, content, and discriminant validity of the measurement, provided compelling evidence for the continued use of the SPPS-T and SPPS-C.
