Functions for Applied Behavioural Sciences in R A general purpose toolbox package in R that consists of a set of useful and varied functions for applied analyses in the behavioural sciences.
Confidence intervals for standardized effect size estimates, for both fixed and random predictors, are generated using randomly constructed distributions and optimized through stochastic approximation. Sample size planning incorporates the uncertainty associated with effect size estimates and provides both quantiles for power as well as expected confidence interval width.
Future functions will include mediation functions that provide p-values for the indirect effect based on the partial posterior p-value and confidence intervals based on hierarchical Bayesian analyses. Functions for resampling regression models are provided for both casewise resampling as well as the wild bootstrap. All regression resampling functions incorporate missing data functionality through multiple imputation.