Individual differences in perceptual information processing and their relation to automobile accident involvement.

Abstract

Used a perceptual-information-processing model of driver decision making as a framework to select and devise predictors of accident involvement. 75 commercial drivers were tested in a driving simulator and were given the Embedded Figures Test, Selective Attention Test, and rod-and-frame test. The predictors of field dependence, selective attention, and complex reaction time significantly related to accident involvement. Initial, simple, and choice reaction time did not relate to accident rate. The visual measures of field dependence and the auditory measure of selective attention were related in the predicted direction, with the field-independent drivers making fewer errors in selective perception. This finding lends support to the importance of the further development of an information-processing model of the driving task. Since evidence is accumulating that stable individual differences in information processing relate to accident involvement, consideration should be given to devising techniques to develop these skills. (16 ref)

Role of premorbid factors in predicting safe return to driving after severe TBI

“…From a theoretical point of view, on roadassessments do not elucidate the strategic level of driving skill that includes all the decisions made before actual driving starts [4,10]. Another difficulty in assuming that on-road tests are valid arises from the consideration that highly-skilled drivers Predictors of post-TBI driving safety 199 sometimes have above-average accident rates: drivers do not always drive as they did during their licensing test [33,60,61]. These arguments undermine the supposed external validity of on-road assessment.…”

“…Various researchers examined fitness to drive in persons with Alzheimer's disease [27,71,72], in subjects with different types of dementia [25,66], in patients with brain damage with or without aphasia [63], in TBI subjects [7,18,20,21,30], in mixed clinical population (typically TBI and cerebrovascular patients together) [11,16,28,29], or in normal subjects [60,73]. Obviously, findings in one population do not necessarily generalize to another.…”

The field dependence/independence cognitive style does not control the spatial perception of elevation

“…This assumption, that there is a processing style that determines how an individual solves such perceptual prob- lems, provides the basis for the conclusion that certain general personality characteristics are developed through chronic activation of a restricted set of field-dependent or field-independent perceptual-processing strategies. Personality characteristics relating to the application of internal/external referents in social situations as diverse as the use of interpersonal space (Witkin, 1977), reactions to the road/driver environment likely to produce an automobile accident (Mihal & Barrett, 1976), and the weighting of others' opinions in conflict situations (Witkin, 1977;Witkin & Goodenough, 1977) were therefore assumed to derive from these perceptual characteristics. Our results, however, are inconsistent with the generality of such a factor of cognitive restructuring in the processing of the egocentric perception of space.…”

Generating multivariate data from nonnormal distributions: Mihal and Barrett revisited

“…Two problems arose in the attempt to do this. First, since the distribution shapes used to model the variables had to be inferred from the descriptive statistics and scale limits reported by Mihal and Barrett (1976), the resulting distributions were approximate rather than exact. Second, when standard normal variables were correlated by the Graybill algorithm and then transformed to follow the desired nonnormal distributions, the intercorrelations were attenuated.…”

“…In the present report, we attempt to solve both of these problems. The actual distribution shapes for the eight variables of Mihal and Barrett's study are ascertained by examining the raw data contained in Appendix A of Mihal's doctoral dissertation (Mihal, 1974), and the attenuation in the rs is eliminated by adjusting the values in the population correlation matrix to compensate for the effects of nonnormality.The computer simulations reported in this paper were conducted on a Macintosh IIci supplied by Apple Computer and the Consortium of Liberal Arts Colleges. The author also wishes to acknowledge the support of NSF-ILl Grant USE-8852l94, awarded to Bates College by the National Science foundation (G. Nigro and D. Bradley, principal investigators.…”