Parsimony in Model Selection: Tools for Assessing Fit Propensity