Overview of our Research
Learning and Memory
Currently, the most active focus of work in our lab is on uncovering methods of enhancing learning and reducing forgetting. Here we look for general principles that have concrete and nonobvious implications for how training and instruction can be optimized. Issues include temporal spacing and novel methods of harnessing the retrieval practice effect. Much of our work is in collaboration with Doug Rohrer of USF. A recent overview of our research is here, and some of the educational implications of the work are discussed here. We are also engaged in collaborative research with Mike Mozer to explore the neurocomputational bases for these effects (also examining human reinforcement learning). Our work is currently supported by the Institute of Education Sciences, the National Science Foundation, and the James S. McDonnell Foundation.
Our lab also has published a great deal of research on multitasking, often using reaction times to dissect fundamental human information-processing bottlenecks, as described in this book and this review article. This work continues, e.g., in a recent paper showing that the central bottleneck found in conventional reaction-time experiments also arises in tasks in which a person makes a free decision in order to optimize an outcome, and a recent Psychological Science paper showing that simulated driving contexts evoke bottleneck effects.
Our lab also published the first demonstration of change blindness (in 1988) and we continue to explore the puzzle of visual awareness and its logical structure-most recently in the Boolean Map Theory of Visual Attention reported in Psychological Review in 2008 with Liqiang Huang (bolstered by empirical results reported in in Science in 2007).
Methodological Issues in fMRI and Cognitive Modeling
Finally, our lab has a keen interest in methodological issues and we have published several widely cited and controversial methodological critiques: one (with Roberts) dealing with the misinterpretation of curve fitting in math modeling, and another in 2009 (with Vul, Harris, and Winkielman): the so-called "Voodoo Correlations paper" revealing a proliferation of statistical error and inflated results in fMRI studies of individual differences (the paper is here, and our reply to commentators is here).