NEWS

PsychoMatic 0.3.0 (2026-06-29)

Updated DESCRIPTION with a more informative title, description, DOI-backed methodological references, maintainer metadata, VignetteBuilder, and LazyData.
Reduced strong dependencies by moving report-only and workflow-specific packages to Suggests and loading them conditionally.
Regenerated English-facing manual pages for exported functions.
Replaced the old NEWS file with NEWS.md.
Added lightweight vignettes for EFA, CFA, measurement invariance, and alignment invariance.
Added small bundled datasets for reproducible examples: psychomatic_continuous, psychomatic_ordinal, and psychomatic_multigroup.
Ensured examples, tests, and report exports use temporary files where appropriate.

Added screen_items() for item-quality screening.
Added reverse_items() for reverse scoring with explicit or detected items.
Added score_scale() for mean/sum scale scores with minimum valid-response rules and optional person-mean imputation.
Added psychomatic_report() for integrated Excel or Word reports.
Added irt_auto() for basic Rasch, 2PL, and graded response IRT workflows, including item-level interpretation, an overall model summary, and APA 7 methodological references.
Added dif_auto() for lightweight logistic DIF screening by group, with item-level interpretation, an overall DIF summary, and APA 7 methodological references.
Added compare_models_auto() for comparing alternative CFA models with fit-index-based recommendations and support for WLSMV-style scaled indices.

Added summary() methods for efa_auto, cfa_auto, and factorial_invariance_auto.
Added more predictable structured tables to EFA, CFA, and invariance result objects.
Preserved bilingual output support through language = "esp" and language = "eng" while keeping the main documentation in English.
Aligned export_efa() with psychomatic_report() so EFA exports use the same structured reporting path as the rest of the package.

Split large EFA and CFA files into focused helpers, methods, exports, and summary utilities.
Kept legacy Spanish aliases for backward compatibility, but documented the English argument names as the recommended interface.

Expanded tests for item screening, reverse scoring, scale scoring, report export, IRT, DIF, CFA model comparison, datasets, summaries, and error paths.
Verified the release candidate with R CMD check --as-cran.

Added inter-item correlation matrices with Pearson, polychoric, and tetrachoric options.
Added Excel export for correlation matrices.