Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
June 20, 2026Open Access

Robust Estimation of 2PL IRT Parameters via Deep Learning under Non-Ideal Testing Conditions

View Full Paper
Ask AI
Bookmark
Share

Authors

YYYunbo Yang

Discussion

Loading...

Member takes

Overview

Randomized trial demonstrates improved parameter estimation in low-resource assessments, suggesting innovative solutions for dynamic testing environments.

Key Points

  • This research aims to enhance the estimation of item response theory parameters under non-ideal testing conditions.
  • Developed a deep learning model based on Dynamic Key-Value Memory Network and Deep-IRT.
  • Compared performance to the Marginal Maximum Likelihood with Expectation-Maximization algorithm.
  • Implemented label smoothing design and modified Pearson residual loss function to improve estimation performance.
  • Deep learning model outperformed MML-EM in terms of Root Mean Squared Error (RMSE) and Pearson correlations.
  • Significant improvements noted under conditions of limited item and examinee sizes, and mismatched distributions.

Cite This Study

Yunbo Yang (2025) studied this question.

synapsesocial.com/papers/6a363224db0793dc1a538c42https://doi.org/10.6082/6d3fp-1dv06
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Likelihood-Free Estimation of IRT Models in Small Samples: A Neural Networks Approach2024 · 4 citations
  2. 2A modified expectation–maximization algorithm for accelerated item response theory model estimation with large datasets2026 · 1 citations
  3. 3IRT Item Parameter Recovery With Marginal Maximum Likelihood Estimation Using Loglinear Smoothing Models2015 · 18 citations
  4. 4The Rank-2PL IRT Models for Forced-Choice Questionnaires: Maximum Marginal Likelihood Estimation with an EM Algorithm2024 · 4 citations
  5. 5Scalable Learning of Item Response Theory Models2024