Sample Size and Fit Statistics in Rasch Models
Author Information
Author(s): Adam B. Smith, Robert Rush, Lesley J. Fallowfield, Galina Velikova, Michael Sharpe
Primary Institution: Cancer Research UK – Clinical Centre, St. James's University Hospital, Leeds, UK
Hypothesis
The study aims to explore the relationship between fit statistics and sample size for polytomous data.
Conclusion
Mean square statistics are relatively independent of sample size for polytomous data, while t-statistics are highly sensitive to sample size.
Supporting Evidence
- The study found that t-statistics identified more misfitting items than mean square statistics.
- Mean square statistics remained stable across varying sample sizes.
- The results suggest that t-statistics may inflate Type I error rates as sample size increases.
Takeaway
This study looked at how the number of people in a study affects the way we check if questions are working well. It found that some methods are better than others when we have a lot of people answering.
Methodology
Data were collected from cancer patients who completed the Patient Health Questionnaire – 9 and the Hospital Anxiety and Depression Scale, with various sample sizes analyzed using Rasch models.
Limitations
The study used real patient data rather than simulated data, which may affect the accuracy of Type I error rates.
Participant Demographics
4072 cancer patients, with 2781 females and 1291 males, average age 60 years.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website