BiasAfterRandomization: Compute weights after randomizing trials or responses

In armanabraham/chb: Choice history biases

Description

This is checker function to ensure that history weights do indeed disapper after trial order is made random or subjects responses are made random. This randomization effectively destroys history effects effectively setting history weights to 0. !!! CHECK IF YOU HAVE ANOTHER FUNCTION THAT DOES THIS!!!! It is a suspect, because uses rawData as input and not glmData, which is much more clean data. Again, compute history weights after either shuffling trial order or shuffling responses to each trial, by either scrambling the order of trials, or by generating responses randomly. This procedure should decrease the history weights to 0. If we run this function many times we will get confidence intervals for history biases

Usage

BiasAfterRandomization(rawData, randomizationType = 1, alpha = 0,
  nBack = 1, B = 10)

armanabraham/chb documentation built on May 10, 2019, 1:39 p.m.