Methods for validating chronometry of computerized tests.
Joshua P Salmon1, Stephanie A H Jones1, Chris P Wright2
1a Department of Psychiatry , Dalhousie University , Halifax , NS , Canada.
Journal of Clinical and Experimental Neuropsychology
|August 18, 2016
Summary
Standard computer hardware can introduce noise, affecting reaction time accuracy. This study presents methods to assess and correct temporal measurement errors in reaction time data.
Area of Science:
- Cognitive psychology
- Psychometrics
- Human-computer interaction
Background:
- Reaction time (RT) is a crucial metric in research and clinical settings.
- Standard computer hardware can introduce measurement error, impacting RT data accuracy.
- Accurate RT measurement is essential for valid research conclusions.
Purpose of the Study:
- To outline methods for assessing the temporal accuracy of reaction time data.
- To demonstrate the detection, measurement, and correction of nonrepresentative noise in RT measurements.
- To provide a cautionary guide for researchers and clinicians using computerized RT assessments.
Main Methods:
- Three methods for assessing temporal accuracy of RT data were outlined.
- One method utilized external chronometry for validation.
- Example data from the Dalhousie Computerized Attention Battery (DalCAB) were used for analysis.
Main Results:
- Standard computer hardware introduces nonrepresentative noise into reaction time measurements.
- This noise can compromise the accuracy and precision of RT data.
- Methods for detecting and correcting such noise were discussed.
Conclusions:
- Researchers and clinicians must assess the temporal accuracy of their RT measurement systems.
- Verifying the internal chronometry of hardware and software is critical for reliable RT data.
- Ignoring potential measurement errors can lead to inaccurate conclusions.
