EQS 6.2AND ADDED FEATURES IN 6.3

WHY USE EQSIRT?

EQSIRT – a comprehensive Item Response Theory computer program – was developed with funding by the U.S. National Cancer Institute through a Small Business Innovation Research (SBIR) grant motivated by NIH’s Patient Reported Outcomes (PROMIS) initiative.

Authors and contributors: EQSIRT was designed by Eric Wu and Peter Bentler, who developed EQS, a popular structural equations modeling program. It was developed with IRT experts Patrick Mair and Steven Reise. Coherent and creative statistical programming was directed by Sam He, and complex interfaces were directed by Guisuo Guo.

EQSIRT has a complete model builder to simplify the building of reusable IRT commands. Like EQS, the program includes a full service data management and general statistics component for preliminary data screening. EQSIRT is transportable across Microsoft Windows, Apple’s Mac OS X (10.6 or higher), and various Linux operating systems.

The methodologies available in eqsirt are the following

Item Response Theory Module

No. 1

If you experience difficulties during installation of the software or are getting a bad disk error, please contact Multivariate Software’s sales department at +1-818-906-0740 or send an email to the sales department at sales@mvsoft.com.

No. 2

Dimensionality Assessment

  • Correlation Based Exploratory Factor Analysis
  • Full Information Exploratory Factor Analysis

No. 3

IRT Models

  1. Unidimensional Models
    • Binary data
      • Rasch, 1PL, 2PL, and 3PL Models
      • Parameter Estimation with MML and MCMC
    • Polytomous Data
      • Graded Response Model (GRM)
      • Generalized Partial Credit Model (GPCM)
      • Nominal Response Model (NRM)
  2. Multidimensional Models
    • Binary data
      • Rasch, 1PL, 2PL, and 3PL Models
      • Multiunidimensional Parameter Estimation with MCMC
    • Polytomous Data
      •  Graded Response Model (GRM)
      • Generalized Partial Credit Model (GPCM)
    •  IRT Models with a Non-Normal Latent Trait (unidimensional only)
    • Models with Equality and Fixed Constraints
    • Multisample IRT Models (both uni and multidimensional models)
    • Latent Traits Computed and Trait Scores Appended in Data Editor
    • IRT Models with Covariates (MML version of GNLMM method)
    • Differential Item Functioning (DIF)
    • Equating
    • Multilevel Models
    • R Scripts Generated for IRT Plots

No. 4

Latent Class Analysis

  • With Combination of Binary, Ordinal, and Continuous Items
  • Estimated Class Membership Appended to Data Editor

No. 5

Mokken Scale Analysis

No. 6

EQSIRT Simulation

Data Management and General Statistical Module

No. 1

Data Import and Export

  • Import Fixed and Free Format Text Data
  •  Import SPSS Data
  •  Import Microsoft Excel Data Sheet
  • Export Data to SPSS
  • Export Data to Excel Data Sheet

No. 2

Data manager

  • Insert and Delete Row and Columns
  • Join Variables and Merge Samples
  • Define Missing Data
  • Sort Cases
  • Comprehensive Case Selections
  • Re-group and Collapse Categories
  • Expand Compacted Data into Regular Data Set
  • Comprehensive Transformation

No. 3

Statistical Analysis (these features are not available in Mac and Linux)

  • Descriptive Statistics, Frequency, and t-Tests
  • Cross-tabulation
  • ANOVA and Generalized Linear Model
  • Regression Analysis
  • Nonparametric Tests
  • Missing Data Diagnosis and Imputation

No. 4

Data Plots on Variable