[Skip to Navigation]
Comment & Response
December 2015

Errors in Derivation Data and Validation Data

Author Affiliations
  • 1Department of Psychology, Georgia State University, Atlanta
  • 2Department of Health Promotion Sciences, Mel and Enid Zuckerman College of Public Health, University of Arizona, Tucson
  • 3Department of Psychology, University of North Carolina at Greensboro, Greensboro
  • 4Institute on Family and Neighborhood Life, Clemson University, Clemson, South Carolina
  • 5Department of Psychology, Wayne State University, Detroit, Michigan
JAMA Pediatr. 2015;169(12):1178-1179. doi:10.1001/jamapediatrics.2015.3076

To the Editor We write to report errors in our article, “Trajectory Analysis of the Campus Serial Rapist Assumption,”1 published online July 13, 2015, in JAMA Pediatrics. After publication of this article, we were asked about the scoring syntax used in the derivation data compared with that used in the publicly available version of the data set. We investigated and found inconsistencies in missing data between the data used for our published analyses and the publicly available derivation data. Consequently, we reconstructed the validation data set and found that 2 eligible cases had been inadvertently excluded from the validation data set used in the published analyses and another 2 ineligible cases inadvertently included. We found no additional errors after fully reconstructing and reanalyzing both data sets. Most of the frequencies and statistics reported in the Results section slightly differ now that the errors are corrected. These differences are due to the small changes in the overall sample size and the 1 additional rapist identified, which produced small changes in the trajectory classification in the validation model. All interpretations and conclusions remain the same after correcting these errors; the 3 trajectory models still best fit the data, and the direction and statistical significance of the coefficients remain unchanged. We have provided a corrected article with corrections to the Abstract, text, Tables and Figure.1

Add or change institution