PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
December 1, 1981Psychometrika2,354 citations

Marginal Maximum Likelihood Estimation of Item Parameters: Application of an EM Algorithm

View Full Paper
RBR. Darrell BockMAMurray Aitkin

Key Points

  • To effectively estimate item parameters using marginal maximum likelihood without making arbitrary assumptions about the ability distribution.
  • Utilized an EM algorithm for maximum likelihood estimation of item parameters in marginal distribution.
  • Characterized ability distribution empirically to avoid assumptions about its form.
  • Applied the procedure to general item-response models with multiple latent dimensions.
  • The EM algorithm effectively estimates item parameters across various item-response models.
  • Demonstrated applicability in models lacking simple sufficient statistics for ability.
  • Successfully handled scenarios with more than one latent dimension.

Abstract

Maximum likelihood estimation of item parameters in the marginal distribution, integrating over the distribution of ability, becomes practical when computing procedures based on an EM algorithm are used. By characterizing the ability distribution empirically, arbitrary assumptions about its form are avoided. The Em procedure is shown to apply to general item-response models lacking simple sufficient statistics for ability. This includes models with more than one latent dimension.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Bock et al. (1981) studied this question.

synapsesocial.com/papers/69d7cfe733ca018b39ae2e83https://doi.org/10.1007/bf02293801
Ask AI
Helpful
Bookmark
Share
View Full Paper