![[PDF]](/pb-assets/Icons/adobe-pdf-icon-1498649896243.png){kind=icon}
Abstract
This paper concerns the problem of estimating the item parameters of latent trait models in a multistage testing design. It is shown that using the Rasch model and conditional maximum likelihood estimates does not lead to solvable estimation equations. It is also shown that marginal maximum likelihood estimation, which assumes a sample of subjects from a population with a specified distribution of ability, will lead to solvable estimation equations, both in the Rasch model and in the Birnbaum model.
References
| Birnbaum, A Lord, FM, Novick, MR Test scores, sufficient statistics and the information structures of testsStatistical theories of mental test scores1968Reading, MAAddison-Wesley Google Scholar | |
| Bishop, YMM, Fienberg, SE, Holland, PW Discrete multivariate analysis: Theory and practice1975Cambridge, MAM.I.T. Press Google Scholar | |
| Bock, RD, Aitkin, M Marginal maximum likelihood estimation of item parameters: An application of an EM algorithmPsychometrika198146443459 Google Scholar | |
| Bock, RD, Lieberman, M Fitting a response model for dichotomously scored itemsPsychometrika197035179197 Google Scholar | |
| Fischer, GH On the existence and uniqueness of maximum likelihood estimates in the Rasch modelPsychometrika1981465977 Google Scholar | |
| Lord, FM Applications of item response theory to practical testing problems1980Hillsdale, NJLawrence Erlbaum Google Scholar | |
| Mislevy, RJ Estimating latent distributionsPsychometrika198449375381 Google Scholar | |
| Rasch, G Probablistic models for some intelligence and attainment tests1960CopenhagenDanish Institute for Educational Research Google Scholar | |
| Wright, BD, Stone, MH Best test design1979ChicagoMESA Press Google Scholar |
