Related Experiment Videos
Data quality in evaluation of an alcohol-related harm prevention program
John W Graham1, Melinda M Roberts, James W Tatterson
1Department of Biobehavioral Health, East 315 Health & Human Development Building, Penn State University, University Park, PA 16802, USA.
Evaluation Review
|April 16, 2002
Summary
This study assessed the reliability and validity of scales measuring substance use and health behaviors in college students. Most scales demonstrated acceptable reliability and validity, supporting their use in future research.
Area of Science:
- Psychology
- Public Health
- Health Behavior Research
Background:
- Substance use and related behaviors are significant public health concerns among college students.
- Accurate measurement of these behaviors and related constructs is crucial for effective prevention and intervention strategies.
- Existing measurement tools require rigorous evaluation for reliability and validity in this population.
Purpose of the Study:
- To evaluate the reliability and convergent validity of 27 composite scales and two items assessing various health-related behaviors and characteristics in college students.
- To provide guidance on data quality assessment and missing data procedures for large, multivariate, longitudinal datasets.
- To establish the psychometric properties of measures related to alcohol, tobacco, marijuana, and other drug use, as well as prevention intentions and actions.
Main Methods:
- A sample of college students completed surveys assessing a range of health behaviors and characteristics.
- Reliability was assessed using internal consistency measures.
- Convergent validity was examined by correlating scale scores with related measures.
- Data quality and missing data procedures were detailed for complex datasets.
Main Results:
- Twenty-three out of 27 composite scales demonstrated at least acceptable reliability.
- The remaining 4 composite scales showed marginally acceptable reliability.
- At least moderate convergent validity was established for 25 of the scales.
- The study illustrated effective data quality assessment and missing data handling techniques.
Conclusions:
- The majority of the assessed scales possess adequate reliability and validity for measuring substance use and related constructs in college students.
- These findings support the use of these scales in research and practice aimed at understanding and addressing health behaviors in this demographic.
- The methodological insights on data quality and missing data are valuable for researchers working with complex longitudinal data.