SNAP judgments: A small N acceptability paradigm (SNAP) for linguistic acceptability judgments: Online Appendices
Kyle Mahowald, Peter Graff, Jeremy Hartman, Edward A. F. Gibson · Language · 2016
STUDY RESULTS'z-bad' is the average z-score for the hypothesized 'bad' option.'z-good' is the average z-score for the hypothesized good option.'Z.diff' is the difference between z-good and z-bad and is the effect size.Beta is the estimate from the linear mixed-effects model, which has a standard error 'SE' and a t-value 't'.'χ 2 ' is the chi-squared value comparing the full model to an intercept-only model, and 'χ 2 p' is the p-value obtained by that comparison.Simple 'p' is just the p-value calculated using the t-value.Pred is TRUE if the effect goes in the significant direction.Sig is TRUE if there is a significant effect.Rows in yellow are rows in which the effect goes in the predicted direction but is not significant.