Methodology for developing and evaluating the PROMIS smoking item banks
Mark Hansen1, Li Cai2, Brian D Stucky3
1CSE/CRESST, Graduate School of Education and Information Studies, University of California, Los Angeles, CA; markhansen@ucla.edu.
Summary
The PROMIS Smoking Initiative developed reliable item banks and short forms for assessing smoking behaviors. These tools enhance measurement accuracy for nicotine dependence and related factors in smoking research.
Area of Science:
- Psychology
- Public Health
- Psychometrics
Background:
- The PROMIS Smoking Initiative aimed to create robust assessment tools for smoking.
- Existing measures often lacked comprehensive evaluation for key smoking constructs.
Purpose of the Study:
- To develop and evaluate item banks, short forms (SFs), and computerized adaptive tests (CATs) for six smoking-related constructs.
- To provide researchers with validated instruments for assessing nicotine dependence, expectancies, and motivations.
Main Methods:
- Utilized response data from a large national sample of smokers.
- Conducted item factor analyses and differential item functioning (DIF) analyses for item selection.
- Calibrated item banks and developed SFs, evaluating performance via simulation studies.
Main Results:
- Developed unidimensional and bias-free item banks through rigorous selection.
- Simulation studies confirmed accurate measurement of constructs using SFs or CATs with few items.
- Demonstrated the efficacy of the developed assessment tools.
Conclusions:
- The PROMIS smoking item banks offer novel tools for measuring smoking constructs.
- Calibrated item banks and SFs improve score estimate quality, advancing smoking research.
- Innovative methods in item selection and SF construction have broader applicability.
