2403 Exercise science researchers rely on statistical methods to evaluate scientific hypotheses and explain complex phenomena. However, misuse of statistical methods results in numerous problems including: inaccurate estimates; poor predictive properties; & type I and type II errors. Assessing goodness-of-fit for well-characterized theoretical distributions to sample data is an important first step in selecting appropriate statistical methods. Yet the growing number of recognized theoretical distributions, as well as newer developments in building application specific models, makes the process of choosing, fitting, and determining goodness-of-fit for different statistical models a complicated task. Furthermore, results obtained from single goodness-of-fit tests are inconclusive, and results obtained from multiple tests are often contradictory. PURPOSE: To propose a strategy for combining the results of formal/quantitative tests with visual/graphical evaluations into an overall goodness-of-fit assessment represented as a point on an evidential support continuum (ESC). METHODS: The theoretical distributions assessed were the Uniform (U[minimum, maximum]), Normal (N[mean, standard deviation]), & Generalized Lambda Distribution (GLD [lambda 1, lambda 2, lambda 3, lambda 4] controlling for location, scale, skewness, & kurtosis, respectively). Data consisted of isotonic knee extension power scores from 24 participants (n = 20 to 40 scores/participant). Quantitative methods included the chi-squared and K-S test procedures. Graphical methods involved assessment of the main body and tails for data histograms vs. distribution probability density functions (p.d.f.s) as well as empirical distribution functions (e.d.f.s) vs. distribution functions (d.f.s). RESULTS: Use of ESCs resulted in the emergence of three distinct classification categories: (1) datasets more suggestive of a particular distribution (N = 13 [n = 3 (U); n = 2 (N); & n = 8 (GLD)]); (2) datasets with sufficient evidence for more than one distribution (N = 9[n = 7 (N & GLD) & n = 2 (U, N, & GLD)]); & (3) datasets with insufficient evidence for any of the three distributions (N = 2). CONCLUSIONS: The ESC approach appears effective as an overall goodness-of-fit assessment. Specifically, ESCs: (a) extend results of formal hypothesis tests (e.g., corroborate conclusive results & clarify ambiguous results); & (b) provide valid and useful inferences of fit quality (e.g., evidential support is described as ‘lesser to greater’ vs. ‘yes or no’ & distributional approximations are considered as ‘adequate, better, or best’ fits overall).