Closed jeanimal closed 5 years ago
See the definitions in G&G 1996 (http://www.dangoldstein.com/papers/FastFrugalPsychReview.pdf). The model we make should match the definition for binary cues. Note this also has a bearing on the issue of handling NA because G&G 1996 define how they do.
Cue reversal is already done for these two models.
G&G 1996 assign 0 weights to NA values so dawesModel and franklinModel should do the same
I am going to close this issue because we implemented some code but then decided it didn't do anything. (So write now there is an un-exported version of tallyModel in the code.)
Probably this means adding the reverse_cues option to dawesModel and franklinModel, TRUE by default but FALSE in the case of tallying.