Rescoring Summary Information

The NIH Toolbox app includes necessary scoring adjustments for five measures within the Cognition Battery. Version 1.8 released on December 15, 2016 contains an automatic rescoring procedure, that will recalculate the scores for data collected prior to this date.   Four measures were adjusted to assure consistency with the original published version (Dimensional Change Card Sort Test, Flanker Inhibitory Control and Attention Test, Pattern Comparison Processing Speed Test, and Picture Sequence Memory Test); one measure (List Sorting Working Memory Test) was adjusted due to a minor scoring error discovered in earlier versions of the iPad app.
A comprehensive study of “mode effects” was recently completed, comparing the web-based version of NIH Toolbox tests, published in 2012, and the current iPad app versions. It is essential to ensure score “equivalency,” as the national norms were derived using the web-based version, so that scores can be assured to have the same meaning. Our study found differences in participant performance between the two versions on the four Cognition instruments noted above for some age ranges, and as a result, we have applied scoring adjustments to the iPad app that result in equivalent scores and normative results. In addition, we discovered a minor scoring error on the List Sorting Working Memory Test that affects a single normative score in the Toolbox app for selected individuals.
All existing scores on this iPad for any of the affected measures and associated composite scores will be automatically updated to reflect the needed adjustments. For the Dimensional Change Card Sort Test, Flanker Inhibitory Control and Attention Test, Pattern Comparison Processing Speed Test, and Picture Sequence Memory Test, this means the Computed Score, Uncorrected Standard Score, Age-Corrected Standard Score, and Fully Corrected T-Score have been updated for all participants. For the List Sorting Working Memory Test, Fully Corrected T-Scores have been updated for some Caucasian and Asian participants. In addition, for all affected measures, participants’ Fluid Cognition Composite and Cognitive Function Composite Scores (uncorrected, age-corrected, fully corrected) will be updated.
Using the app’s export function will initiate this scoring update on the device, and all affected assessments will then reflect corrected scores for the five Cognition measures. These changes assure that score interpretations for these affected measures, as well as associated composite scores, reflect U.S. census-weighted normative values as developed from the NIH Toolbox national norming study. In addition, going forward, any data collected using this version of the app (and future versions) will provide accurate and equivalent scores.  An audit log will automatically be created to identify all the records that have been changed and where those changes occurred.  The audit log can be accessed at the top of this screen and can be exported via email, iCloud or iTunes.  The audit log is not designed to be used in analysis, but as a record of the changes that were made during rescoring.  When reviewing the audit log please note that you will see BOTH the old data and the new rescored data.  The records will be paired together with the new (v2.1) above the old (v2.0).  The only exception being the rescored composite scores, where you will only find the new records (v1.1) in the audit log.   


Did you find this article helpful?