Umass Psychology People
Dr. Alexander Pollatsek
Position(s): Research Faculty
My research in the last 5-10 years has mainly focused on using eye movements to study reading and driving. In reading, my work has largely been on how people process words in the context of normal reading rather than in laboratory tasks such as “lexical decision”. More specifically, a lot of my work has been on how people process morphemically complex words such as compound words and prefixed and suffixed words. Much of the data indicates that the standard models of word recognition, that were developed for short simple words, are inadequate for these more complex words. I have also been continually involved in working on developing our model of eye movements in reading, the E-Z Reader Model, both to deal with more complex reading situations and to understand whether eye movement control is fundamentally different in reading than in scene recognition and visual search.
My research on driving, done in collaboration with Donald Fisher of the department of Mechanical and Industrial Engineering, uses eye movements to study issues related to driver safety. One of our major projects has been on the anticipation of potential hidden hazards. Failure to anticipate such hazards has been shown to be both a major cause of accidents and a major cause of an increased accident rate for novice drivers (relative to more experienced drivers). We have shown, both in a driving simulator and on the road, that there are indeed large differences between these groups and we have developed a training program on a PC that takes less than an hour to complete that makes novice drivers about as good as experienced drivers. This program, and variants of it, is now starting to be widely used in driver training. A second major cause of accidents is failure to attend to the forward roadway. We have also established that there are large differences between novice drivers and more experienced drivers on this failure and have developed a second training program for this. Our preliminary results for this training program are also encouraging.
I have not been as active in the last 5 years on research in understanding statistical concepts (done in collaboration with Clifford Konold of the Scientific Reasoning Research Institute). One of our concerns is that a major problem in understanding statistics is that people don’t really understand what a mean or average is beyond knowing what the computational formula is. However, we are planning to embark on research in this area, using among other techniques, eye movements to study whether novices and experts examine histograms differently when they need to examine two histograms to decide whether two groups differ on the dimension represented in the histogram.