A Comparison of Several Goodness-of-Fit Statistics
- 1 March 1985
- journal article
- research article
- Published by SAGE Publications in Applied Psychological Measurement
- Vol. 9 (1), 49-57
- https://doi.org/10.1177/014662168500900105
Abstract
A study was conducted to evaluate four goodness- of-fit procedures using data simulation techniques. The procedures were evaluated using data generated ac cording to three different item response theory models and a factor analytic model. Three different distribu tions of ability were used, as were three different sam ple sizes. It was concluded that the likelihood ratio chi-square procedure yielded the fewest erroneous re jections of the hypothesis of fit, whereas Bock's chi- square procedure yielded the fewest erroneous accep tances of fit. It was found that sample sizes some where between 500 and 1,000 were best. Shifts in the mean of the ability distribution were found to cause minor fluctuations, but they did not appear to be a major issue.Keywords
This publication has 7 references indexed in Scilit:
- Testing the conditional independence and monotonicity assumptions of item response theoryPsychometrika, 1984
- Marginal maximum likelihood estimation of item parameters: Application of an EM algorithmPsychometrika, 1981
- Using Simulation Results to Choose a Latent Trait ModelApplied Psychological Measurement, 1981
- When are item response models consistent with observed data?Psychometrika, 1981
- A goodness of fit test for the rasch modelPsychometrika, 1973
- Estimating item parameters and latent ability when responses are scored in two or more nominal categoriesPsychometrika, 1972
- Generating multiple samples of multivariate data with arbitrary population parametersPsychometrika, 1965