**Getting More from Visual Working Memory:
Retro-Cues Enhance Retrieval and Protect from Test Interference**
Alessadra S. Souza, Laura Rerko, & Klaus Oberauer
Paper submitted for publication at JEP:HPP
--------------------------------------------------------------------
This page documents the data of the experiments reported, the analyses performed, and the results thereof.
List of Experiments (E):
E1 and E2 use color recognition tasks
E3 - E6 use continuous color reproduction tasks
For E1 & E2, the data was analyzed with an R script, looking at response accuracies. Statistical analysis was performaned with the BayesFactor package.
For E3 to E6, we first looked at the distribution of response deviations (raw data) using an R script.
Next, we submitted the data to mixture modeling (MatLab script) to estimate parameters related to the maintenance of memory representations:
(1) precision with which objects were stored in memory (sigma)
(2) probability of recalling one of the non-target objects (beta)
(3) probability of guessing with a random color (gamma)
and probability of recalling the target object (target) computed as 1 - (beta + gamma). This was not a free parameter in the model!
The mixture model follows the model proposed by Bays, et al. (2009).
We extended this mixture model to include one further component related to color wheel interference
(4) the wheel attraction effect (cwi)
We estimated the parameters by fitting the models for each individual in each condition separately. We compared estimated parameters across conditions using Bayesian analyses.
