changes: *added error checks for common issues - too few trials and missing conditions
changes: * added examples in the function help and improved function documentation * added validation vignette (tests many combinations of options within splithalf functions, incluing the multiverse functions) * fixed minor bugs in some combinations of options within splithalf (typically within the ) * fixed some documentation * fixed plotting issue with score = “difference_of_difference” * minor improvements to documentation
changes: * minor updates: reworked documentation and README * added hexlogo * bug fixes to splithalf.multiverse
changes: * added multiverse analyses to package. Functions splithalf.multiverse, testretest.multiverse, plot.multiverse, and threshold now available!
changes: * added “sum” option to the average input. average = “sum” can now be used to calculate the sum score for accuracy (binary 0/1 values)
changes: * added DPrime option, this is still under development * the structure of splithalf has changed to include additional elements in the output list - intended to make the future reliability multiverse add-on easier to use * added a basic plotting functionality, setting plot = TRUE will now give a raincloud plot of all estimates (based on code from Allen et al. https://wellcomeopenresearch.org/articles/4-63/v1) * added a ‘round.to’ option to specify the number of decimal points to round the estimates to
changes: * updated code to allow for estimation of splithalf reliability of accuracy rates. (beta testing needed) * minor bug checks, fixed using spearman over pearson correlation.
changes: * a guide to the package can be found here https://sdparsons.github.io/splithalf_documentation/ * bug fix to the spearman-brown correction, it now deals with negative splithalf estimates properly. * fixed error in splitting the sample in the halfs type of split (Thanks to Anne-Wil Kruijt for pointing out) * updated documentation
Changes: * all functions compressed into a single function “splithalf” * added utilities such as RT trimming have been removed, the function requires you use data that has been processed * added C++ code with Rcpp to massively speed up the function.
I forgot to record changes because I am awful - lots of improvements mainly to speed things up
Changes: * added dplyr to dependencies * added splithalf functions for accuracy (currently development version)
Changes:
Changes:
Changes: