New Adaptive Testing Algorithm for Health Literacy Assessments
Author Information
Author(s): Kandula Sasikiran, Ancker Jessica S, Kaufman David R, Currie Leanne M, Zeng-Treitler Qing
Primary Institution: Department of Biomedical Informatics, University of Utah
Hypothesis
Can a new classification method for developing brief health literacy assessment instruments be created without pretesting on large participant populations?
Conclusion
The MDT-based approach is a promising alternative for computerized adaptive testing in health literacy assessments.
Supporting Evidence
- The algorithm classified 88.5% of subjects correctly in the familiarity data set.
- The average test length was reduced by about 50% in the familiarity data set.
- In the numeracy data set, 96.9% of subjects were correctly classified with a 35.7% reduction in test length.
- The method can be validated with few subjects and test items.
Takeaway
This study created a new way to test how well people understand health information, making it quicker and easier for them.
Methodology
The study developed an algorithm using measurement decision theory and applied it to two health literacy data sets.
Potential Biases
The calibration sample may not represent the overall population, affecting accuracy.
Limitations
The study relies on secondary data and assumes independence of responses, which may not hold true.
Participant Demographics
Participants included adults from various age groups and educational backgrounds, with a focus on those with low health literacy.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website