Errors in Items and Algorithm in Questionnaire for Validation Study

To the Editor We write to report a scoring algorithm error that was included in our article, "Development and Initial Validation of a Consumer Questionnaire to Predict the Presence of Ear Disease," published online on August 3, 2017, and in the October 2017 issue of JAMA Otolaryngology–Head & Neck Surgery. In this study, we examined the validity of the Consumer Ear Disease Risk Assessment (CEDRA), a questionnaire for the self-assessment of risk of ear diseases associated with hearing loss. The error was discovered by a member of our research team who was preparing a digital version of the questionnaire used in our study. The error occurred in the analysis code for 3 of the 28 questionnaire items included in the scoring (questions 16, 18, and 19). For these 3 questions, the 4-item rating scales were dichotomized incorrectly. After correcting the algorithm for these questions, we found slightly different results, without any differences in statistical significance of these results. The overall results and conclusions remain unchanged.

http://ift.tt/2FzlwIO