In contrast to an extended literature comparing older and younger drivers, limited direct attention has been given to the safety of women as drivers relative to men. As well as reviewing material bearing on this latter issue, the commentary is a technical report that identifies conceptual and methodological issues that need to be addressed in comparing men and women as drivers.
Limited direct consideration has been given to the relative safety of men and women as drivers. Although sex differences have sometimes been noted through secondary analysis of measures, studies have often reported data for men and women combined. Where there have been explicit attempts to identify relative risk for men and women, the analyses typically have been no more than descriptive in orientation. Even at a descriptive level there are substantial methodological issues that need to be addressed in determining whether men and women differ in safety as drivers (or if relative driver safety has changed over time). Additional requirements bearing on validity of inference need to be satisfied if data on men and women as drivers are to be interpreted in theoretical or explanatory terms.
Research to date on sex differences in driving is deficient on methodological and conceptual grounds. When comparing men and women it is important to keep in mind that sex as a variable is inevitably confounded with many processes that potentially impact on the outcome of Interest.
The question then arises as to whether differences obtained in outcome between men and women are more appropriately attributed to the confounds than described as sex differences. As an example, there are driving exposure differences between men and women. Since crash risk varies with exposure, the issue is whether crash differences between men and women reflect anything more than exposure differences. One strategy has been to adjust crash statistics by making allowance for exposure. Hence fatality rates are generally expressed with reference to distances driven. Even exposure based risk estimates can be confounded by a number of factors such as driver age, trauma consequence, driver experience, vehicle characteristics, and driving habits.
Type: Research and Analysis Report
Sub Type: Consultant Report
Author(s): R Over
ISBN: 0 642 25519 9
ISSN: 0810-770X
Topics: Gender, Methodology
Publication Date: 01/05/98