This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by Arthur, W. Articles by Tubre, A. H. Search for Related Content PubMed Articles by Arthur, W., Jr. Articles by Tubre, A. H. Social Bookmarking                         What's this?

Convergence of Self-Report and Archival Crash Involvement Data: A Two-Year Longitudinal Follow-Up

Texas A&M University, College Station, Texas

Suzanne T. Bell

Texas A&M University, College Station, Texas

Bryan D. Edwards

Texas A&M University, College Station, Texas

Eric Anthony Day

University of Oklahoma, Norman, Oklahoma

Travis C. Tubre

Texas A&M University, College Station, Texas

Amber H. Tubre

Texas A&M University, College Station, Texas

This study constructively extends Arthur et al. (2001) by assessing the convergence of self-report and archival motor vehicle crash involvement and moving violations data in a 2-year longitudinal follow-up. The relationships among these criteria, conscientiousness, and driving speed were also assessed using both predictive and postdictive criterion-related validation designs. Data were collected from a 2-year follow-up sample of 334 participants. Results suggested a lack of convergence between self-report and archival data at both Time 1 and Time 2. In addition, the predictor/criterion relationships varied across research design and data source. An actual application of our findings is that the interpretation of relationships between specified predictors and crash involvement and moving violations must be made within the context of the criterion-related validation design and criterion data source. Specifically, predictive designs may produce results different from those of postdictive designs (which are more commonly used). Furthermore, self-report data appear to include a broader range of incidents (more crashes and tickets), and thus researchers should consider using self-report data when they are interested in including lower threshold crashes and tickets that may not be reported on state records (e.g., because of the completion of a defensive driving course)

Human Factors: The Journal of the Human Factors and Ergonomics Society, Vol. 47, No. 2, 303-313 (2005)
DOI: 10.1518/0018720054679416


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?