Questionnaire discrimination: (re)-introducing coefficient δ
2007

Improving Questionnaire Discrimination with Coefficient δ

Sample size: 14761 publication Evidence: moderate

Author Information

Author(s): Matthew Hankins

Primary Institution: King's College London, Department of Psychology

Hypothesis

Can a new coefficient δG improve the discrimination ability of questionnaires in clinical research?

Conclusion

The new coefficient δG effectively distinguishes between the discrimination capabilities of different scoring methods for questionnaires.

Supporting Evidence

  • The reliability of both scoring methods was high, with Cronbach's alpha above 0.88.
  • Dichotomous coding resulted in a lower discrimination index (δG = 0.73) compared to polytomous coding (δG = 0.96).
  • More than half of the sample scored the same on the dichotomous scale, indicating poor discrimination.

Takeaway

This study shows that using a new method to score questionnaires can help tell the difference between people’s health better than older methods.

Methodology

The study derived a general formula for coefficient δG and applied it to data from the GHQ-12 questionnaire.

Limitations

The study primarily focuses on the GHQ-12 and may not generalize to all types of questionnaires.

Participant Demographics

Participants were from the general population in the UK.

Statistical Information

P-Value

< 0.001

Confidence Interval

95% CL: 0.956, 0.959

Statistical Significance

p < 0.001

Digital Object Identifier (DOI)

10.1186/1471-2288-7-19

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication