summary: S3 Methods: summary

In ForceChoice: Forced-Choice Modeling Based on Item Response Theory and Cognitive Diagnostic Models

S3 Methods: summary

Description

Generates structured, comprehensive summaries of objects produced by the ForceChoice package. This generic function dispatches to class-specific methods that extract and organize key information including model configurations, fit statistics, parameter estimates, and convergence diagnostics. Designed for programmatic access and downstream reporting.

Usage

## S3 method for class 'good.of.fit'
summary(object, digits = 4, ...)

## S3 method for class 'MIRT'
summary(object, digits = 4, ...)

## S3 method for class 'MGPCM'
summary(object, digits = 4, ...)

## S3 method for class 'MGGUM'
summary(object, digits = 4, ...)

## S3 method for class 'FCMIRT'
summary(object, digits = 4, ...)

## S3 method for class 'FCDCM'
summary(object, digits = 4, ...)

## S3 method for class 'FCGDINA'
summary(object, digits = 4, ...)

## S3 method for class 'FCGGUM'
summary(object, digits = 4, ...)

## S3 method for class 'TIRT'
summary(object, digits = 4, ...)

Arguments

object	An object of one of the following classes: Fitted model objects: "MIRT", "MGPCM", "MGGUM", "FCMIRT", "FCDCM", "FCGGUM", "FCGDINA", "TIRT" Goodness-of-fit objects: "good.of.fit"
digits	Number of decimal places for numeric output (default: 4). Applied uniformly across all methods.
...	Additional arguments passed to or from other methods (currently ignored).

Details

Each method returns a named list with class-specific components optimized for structured access and formatted printing:

Fitted Model Objects

Returns a summary.<Class> object with components:

call

Original function call.

model.info

List: family, description, D, method, and model-specific fields (model, fc.type, dcm.type, I.states, N.block).

data.info

List: N, I or B, response.type.

fit.stats

List: LogLik, npar, AIC, BIC.

par.summary

Matrix: parameter summary (Min, Q1, Median, Mean, Q3, Max) across items/statements/blocks for each parameter type.

theta.summary

Matrix: person parameter summary per dimension.

Corr

Matrix: factor inter-trait correlation matrix (where applicable).

convergence

List: algorithm, batch counts, convergence flags, latent grid length L (iStEM) or diagnostic note (Stan).

digits

Numeric: precision used for formatting.

Additional model-specific components:

delta.summary

FCDCM only: higher-order delta parameters.

classification

FCDCM only: observed attribute patterns, mean maximum posterior.

lambda.summary

TIRT only: factor loading summary.

Goodness-of-Fit Objects (good.of.fit)

Returns a summary.good.of.fit object with components:

Sample_Size, Items, Response_Type, Parameters, Moments, LogLik, Deviance

Overview statistics.

IC

Named vector: AIC, AICc, BIC, CAIC, SABIC, HQIC.

M2

List: M2 statistic, df, p-value, RMSEA with CI.

Residual_Fit

Named vector: SRMSR, RMSR, max standardized residual.

Comparative_Fit

Named vector: CFI, TLI, IFI.

Pseudo_R2

Named vector: McFadden, Cox–Snell, Nagelkerke, Aldrich–Nelson, Veall–Zimmermann.

Local_Dependence

Named vector: Q3 mean, max, adjusted max, P95.

Classification

Named vector: posterior entropy, max posterior.

Null_Model

Named vector: null model LogLik, deviance, M2, df.

tables

List of data frames for formatted printing.

Value

Invisibly returns a structured list containing summary components. The exact structure depends on the class of object. All returned objects carry an appropriate S3 class (e.g., "summary.MIRT", "summary.good.of.fit") for use with corresponding print methods.

Methods (by class)

• summary(good.of.fit): Summary method for good.of.fit objects. Extracts and structures all fit indices: information criteria, limited-information M2, RMSEA, SRMSR, CFI/TLI/IFI, pseudo-R^2, local dependence (Yen's Q3), posterior classification, and null-model diagnostics.

• summary(MIRT): Summary method for MIRT objects. Multidimensional IRT (1PL–4PL): extracts binary-response model configuration, fit statistics (LogLik, AIC, BIC), factor correlation matrix, person and item parameter summaries, and convergence diagnostics.

• summary(MGPCM): Summary method for MGPCM objects. Multidimensional Generalized Partial Credit Model: polytomous responses with category-specific step parameters.

• summary(MGGUM): Summary method for MGGUM objects. Multidimensional Generalized Graded Unfolding Model: ideal-point polytomous responses with discrimination (a), location (delta), and threshold (tau) parameters.

• summary(FCMIRT): Summary method for FCMIRT objects. Forced-Choice Multidimensional IRT: dominance model with sequential ranking over item endorsement logits at the block level. Extracts block configuration, FC type (RANK/MOLE/PICK), and statement-level parameter summaries.

• summary(FCDCM): Summary method for FCDCM objects. Forced-Choice Diagnostic Classification Model: higher-order trait with exact marginalization over 2^D attribute profiles. Extracts DCM type (DINA/DINO), higher-order delta parameters, block eta parameters, and posterior classification diagnostics.

• summary(FCGDINA): Summary method for FCGDINA objects. Forced-Choice GDINA model: CDM item model (DINA/DINO/ACDM/GDINA) with sequential forced-choice ranking over item endorsement logits.

• summary(FCGGUM): Summary method for FCGGUM objects. Forced-Choice Generalized Graded Unfolding Model: ideal-point model with sequential ranking over binary GGUM endorsement logits. Extracts forced-choice block structure, unfolding parameters (a, delta, tau), and convergence info.

• summary(TIRT): Summary method for TIRT objects. Thurstonian IRT for Forced-Choice: pairwise probit comparison of latent utility differences. Extracts statement loadings (lambda), uniquenesses (psi2), pairwise gamma matrix, and factor correlations.

ForceChoice documentation built on Sept. 13, 2026, 1:06 a.m.