Validity testing of a questionnaire_PG class
- 1. Questionnaire Validity and Its Testing For PG Students in Public Health Department of Community Medicine
- 2. 1. What is Validity? • Validity: Does the questionnaire measure what it's supposed to measure? • • Ensures accurate, meaningful, and useful data • • Essential for reliable conclusions in public health research
- 3. 2. Types of Validity • Face Validity • • Content Validity • • Construct Validity • • Criterion Validity
- 4. 3. Face Validity • Superficial judgment by experts or stakeholders • • Ask: 'Does it look like it measures the intended concept?' • • Used during initial questionnaire design
- 5. 4. Content Validity • Ensures all aspects of the concept are covered • • Assessed by expert panel using relevance ratings • • Key tools: I-CVI (Item-level), S-CVI (Scale- level) • • Acceptable I-CVI ≥ 0.78; S-CVI ≥ 0.90
- 6. 5. Construct Validity • Assesses if questionnaire truly measures theoretical construct • • Common in attitude, belief, and behavior scales • • Methods: Factor Analysis, Hypothesis Testing
- 7. 6. Criterion Validity • Compares questionnaire to external standard or outcome • • Types: Concurrent Validity, Predictive Validity • • Used when gold standard or outcome is available
- 8. 7. Statistical Tools for Validation • CVI (I-CVI, S-CVI): Content validity • • Kappa statistic: Agreement beyond chance • • Cronbach’s Alpha: Internal consistency • • Factor Analysis: Underlying constructs • • Correlation/Regression: Criterion validity
- 9. 8. Example: Junk Food Questionnaire • 10 Likert scale items on junk food behavior • • Face Validity: Reviewed by faculty • • Content Validity: S-CVI = 0.92 from 4 experts • • Construct Validity: Cronbach’s alpha = 0.82 • • Criterion Validity: Compared with 24-hr recall
- 10. 9. Tips for Students • Define your construct clearly • • Use simple, understandable language • • Pre-test your tool before formal validation • • Document validation steps in your thesis • • Use both qualitative and quantitative methods
- 11. 10. Summary Table • Face Validity: Expert opinion • • Content Validity: I-CVI, S-CVI from expert ratings • • Construct Validity: Cronbach’s alpha, factor analysis • • Criterion Validity: Correlation with standard/outcome