Appropriateness and efficiency of combined response speed and accuracy measures
Cooperation project with Dr. Heinrich René Liesefeld (LMU Munich)
In many psychological experiments, response speed and response accuracy data are gathered. In the ideal case, experimental effects go into the same direction in both measures (or only in one measure while the other is unaffected). In practice, this is, however, sometimes not the case and the data exhibit a speed-accuracy tradeoff (SAT), as exemplified in the right visualization. This situation often creates problems for a straightforward interpretation of either measure.
In the past, several methods were suggested to combine speed and accuracy in this case, but the appropriateness of doing so is still a matter of debate. In this project, we assess the behavior of existing measures (and a newly developed one) by means of simulation studies in an attempt to clarify the appropriateness and justification to use such combined measures.
Liesefeld, H.R., & Janczyk, M. (2019). Combining speed and accuracy to control for speed-accuracy tradeoffs (?). Behavior Research Methods, 51, 40-60.
In some of our published research, we have also employed the drift-diffusion model to investigate specific questions (see, e.g., Janczyk, Mittelstädt, & Wienrich, 2018). We are currently extending this line of research in several ways: (1) We are developing a leaky competing accumulator (LCA) account for dual-task costs. (2) We are developing a PDE solution for drift-diffusion models where the drift rate is not constant but time-dependent. This research is done in collaboration with Prof. Dr. Thomas Richter (University of Magdeburg).
Publications on Statistics and Methods
We have also published articles and tutorials on how to use statistical tools, such as confidence intervals (e.g., Pfister & Janczyk, 2013) and also written a Springer textbook on inferential statistics which was published in the 2nd edition already (Janczyk & Pfister, 2015)