Skip to contents

Changelog

Source: NEWS.md

EFAtools 1.1.0.9000

  • efa_fit() changes:

    • The default rotation is now oblimin (rotation = "oblimin"). With one factor, no rotation is done. efa_group() and efa_mi() use the same default; for this oblique default, efa_group() aligns the groups to the first group unless reference_group is given. EFA() and EFA_POOLED() keep no rotation as their default.
    • PAF now gives the same chi-square-based fit indices as ULS (chi-square, CFI, TLI, RMSEA, AIC, BIC, ECVI), because a converged PAF solution is the ULS solution (Jöreskog & van Thillo, 1971). Thus, PAF without N now also gives a warning that these indices are not available. efa_average() and efa_mi() report them for PAF too.
    • print() shows the chi-square, CFI, TLI, RMSEA, and SRMR. summary() adds AIC, BIC, and ECVI. CAF is not printed, because a correctly specified model gives a CAF near 0.5, not near 1. A CAF that cannot be computed is now NA, not 0, and gives no warning.
    • With se = "np-boot", the bootstrap intervals of RMSEA, CFI, and TLI now come from cases transformed so that their misfit equals the estimated misfit (Yuan, Hayashi, & Yanagihara, 2007). The earlier intervals moved toward worse fit, because each resample added sampling misfit. Their coverage was often much too low (Zhang & Savalei, 2016). print() and summary() show the intervals of CFI and TLI. The other fit indices no longer get bootstrap standard errors or intervals. The new intervals need cor_method = "pearson", no missing values in the analysed cases, and a chi-square with df > 0; otherwise, SE, CI, and replicates have no fit_indices. The additional fits make the bootstrap slower. In efa_mi(), fit_indices_descriptive now pools these replicates of RMSEA, CFI, and TLI.
    • With pairwise deletion of missing values, a new warning (efa_pairwise_n) shows that N, the number of rows, overstates the cases behind each correlation, so the chi-square is too large and the standard errors are too small. cor_method = "fiml" (or efa_mi() for ordinal data) avoids this.
    • se = "np-boot" with a correlation matrix now gives an error (efa_boot_cormat), as the other unavailable standard-error methods do.
    • efa_fit() no longer has .... Give the rotation-engine arguments gam, delta, and maxit to rotate_control().
    • Raw data no longer give the message that correlations are computed from the raw data.
    • The parallel workers of the bootstrap now also compute the correlations of the resamples, which take most of the time for polychoric and FIML correlations. A parallel plan is thus much faster there. This change alone leaves the results for a given seed unchanged.
    • A fit with se = "sandwich" no longer returns Gamma, which needs about 11 MB at 50 variables.
    • The printed header shows N and a correlation type other than Pearson, and the Heywood alert names the variables. A just-identified model (df = 0) shows only SRMR and df.
    • The tables of analytic confidence intervals name their covariance (expected information, robust sandwich, or two-stage FIML sandwich).
    • summary() no longer repeats the simple-structure counts and the largest residual in the model diagnostics; it states when no simple-structure problem occurs. summary(ci = "separate") was removed, because it did the same as "auto".
    • The message about the rows dropped for an asymptotic covariance now gives their number.

  • A tuning knob set in estimate_control() or rotate_control() together with a type preset no longer gives a warning (efa_type_override). The value you set is used.

  • Removed unnecessary warning from efa_map() that warned when the residual variance reached 0 and thus the search ended early.

  • efa_kgc() now counts the eigenvalues greater than 0, not greater than 1, for eigen_type = "SMC" and "EFA". These eigenvalues come from the reduced correlation matrix, and the cutoff 1 has no basis for it. With SMCs, the number of positive eigenvalues is a lower bound to the number of factors (Guttman, 1954). These types now suggest more factors, often too many in samples. "PCA" keeps the cutoff 1. The report and the plot show the cutoff of each type. KGC() and N_FACTORS(criteria = "KGC") change too, and so does efa_retain() with eigen_type_other = "SMC" or "EFA".

  • efa_hull() no longer gives a warning when it increases the largest number of factors that it fits to 3. This is a rule of the method, not a change to a user setting.

  • efa_nest() and efa_parallel() are much faster for large samples, and so are efa_hull() and efa_retain(), which use them. They now draw each simulated correlation matrix directly from its exact distribution and do not simulate raw data (efa_parallel() does this for Pearson correlations only). The reference eigenvalues have the same distribution as before, but the results for a given seed are different.

  • efa_retain() changes:

    • Parallel analysis, also in the Hull method, now uses PCA eigenvalues by default (eigen_type_other = "PCA", eigen_type_HULL = "PCA"), not SMC, which often retains too many factors (Lim & Jahng, 2019). The Kaiser-Guttman criterion and the scree plot also use PCA eigenvalues. For the upper bound of the Hull method, parallel analysis now uses the 95th percentile (new argument decision_rule_HULL = "percentile"; Lorenzo-Seva et al., 2011); decision_rule applies only to PARALLEL, and N_FACTORS() keeps the mean rule for both.
    • The Hull method now uses only the CAF by default (gof = "CAF"). With strongly correlated factors, CFI and RMSEA can make the Hull method retain too few factors. The CAF works with every estimator, so estimator = "PAF" no longer gives a message.
    • Because of these new defaults, the results can be different from earlier versions. N_FACTORS() keeps the earlier defaults.
    • With cor_method = "poly" or "tetra", CD now runs on the Pearson correlations of the raw data. This is the original method (Ruscio & Roche, 2012): its comparison data resample the observed values. Bartlett’s test also uses the Pearson correlations of the raw data, because its chi-square distribution is not valid for polychoric correlations.
    • Fewer warnings. Skipped criteria give one warning, and failed criteria give one more. A default criterion that cannot use the input is skipped without a warning; a missing N still gives one. A skipped Bartlett’s test shows only in the report.
    • An invalid argument (for example n_datasets = 0) now gives an error before any criterion runs, not a warning that a criterion could not be run.
    • The Hull method now uses the simulated eigenvalues of the parallel analysis if both run with the same type of eigenvalues, so the simulation runs only once.
    • The suitability report is shorter and shows the same lines as efa_screen().
    • alpha is renamed to alpha_CD, which you must give by its full name. alpha is deprecated.

  • efa_hull() now has the same defaults as efa_retain(): the ML estimator, only the CAF (gof = "CAF"), and PCA eigenvalues with the 95th percentile in the parallel analysis that sets the largest number of factors to fit (eigen_type = "PCA", decision_rule = "percentile"; Lorenzo-Seva et al., 2011). Thus, the results can be different from earlier versions. HULL() keeps the earlier defaults (PAF, all three indices, SMC eigenvalues, and the mean rule).

  • efa_power(mode = "simulation") uses the new efa_retain() defaults for KGC, PARALLEL, and HULL, so their variants are now KGC_PCA, PARALLEL_PCA, and HULL_CAF.

  • efa_screen() changes:

    • Outliers are now detected with the deterministic DetMCD algorithm, so the result is the same on every run and large samples are much faster. The seed argument is deprecated and has no effect.
    • mcd_alpha now defaults to 0.75 instead of 0.5, which flags fewer clean cases.
    • If any variable is categorical (fewer than ten distinct values), no case is flagged as an outlier. The report names the cases with the largest Mahalanobis distances instead. A chi-square cutoff is not valid for rating-scale data, and it flagged many clean cases when items were skewed.
    • The Henze-Zirkler test was removed; Mardia’s tests remain and are much faster in large samples.
    • With cor_method = "poly" or "tetra", the multivariate normality tests are skipped, and the report no longer recommends polychoric correlations for items with few categories.
    • A non-significant Bartlett’s test now appears in the recommendations, instead of the note that the data appear suitable.
    • The multicollinearity recommendation names the variables involved (Belsley’s variance-decomposition proportions), and the report shows the condition index instead of the determinant and the condition number.
    • The report is shorter, and conditions that the report already describes no longer also give a warning.

  • efa_reliability() changes:

    • An efa_fit() solution with an orthogonal rotation is now scored like an oblique solution, with factor correlations of 0, and a "bifactorT" solution as a bifactor solution. Before, the function refused both. The function now refuses a "bifactorQ" solution (efa_reliability_bifactor_oblique), because its group factors are correlated; before, it scored the general factor as a group factor. It also refuses an unrotated solution with two or more factors (efa_reliability_not_rotated).
    • Without factor_map, a loading matrix and the components use their non-zero pattern only if most variables have an exact zero group loading, as in a target matrix. Otherwise, each variable is assigned to the group factor with its largest loading. Before, a loading matrix always used its non-zero pattern, so an estimated bifactor matrix assigned every variable to every group factor; a pattern matrix with Phi and the components always used the largest loading.
    • A lavaan fit is now a bifactor solution only if variables load on a factor named g_name and on other factors. The function now scores correlated-factors fits with cross-loadings and ESEM fits; before, it refused them. In an ESEM block, each variable is assigned to the factor with its largest loading. OMEGA() refuses such fits with efa_reliability_invalid_lavaan.
    • The whole-scale omega subscale now counts all group-factor loadings also with variance = "correlation". Thus, on the whole-scale row, omega total is omega hierarchical plus omega subscale with both settings. Before, it counted only the loadings of the assigned variables, as psych::omega() does. OMEGA() changes too.
    • A correlated-factors solution no longer gives omega subscale. Omega subscale is the share in addition to a general factor, and when the factors correlate, it is not a share of the variance (Rodriguez et al., 2016). For a lavaan fit to a correlation matrix, the omega total of each factor equals semTools::compRelSEM(fit, obs.var = FALSE). Negatively correlated factors no longer give a false out-of-range warning.
    • The function no longer gives messages about a single factor, a correlated-factors lavaan fit, a second-order fit, and variables with fewer than two loadings. print() no longer shows the total-variance line and the note about the omega columns of a correlated-factors solution.
    • A non-positive residual variance of a variable in a lavaan fit now gives the Heywood warning of the other inputs, not an error. A non-positive variance or residual variance of a factor still gives an error, with the new class efa_reliability_factor_heywood. OMEGA() changes too, and now gives the Heywood warning also with add_ind = FALSE.
    • The result has a new attribute, factor_map: the assignment of the variables to the group factors that was used.
    • The mixed-keying warning now also tells what to do when the factors measure opposite constructs. fac_names given with a lavaan fit now gives the efa_reliability_args_ignored warning.
    • g_load, s_load, var_names, and pattern are deprecated. Give the loadings as one matrix in model instead.
    • With the largest-loading rule, a variable without a group loading is no longer assigned to the first group factor. OMEGA() changes too.
    • Without factor_map, a Schmid-Leiman solution in which most variables have an exact zero group loading, such as a transformed lavaan solution, now uses its non-zero pattern, as the lavaan fit does.
    • A lavaan fit with a negative path between latent variables, in addition to the second-order loadings, now gives the error efa_reliability_higher_order. Before, it was scored as a second-order model without that path. OMEGA() changes too.

  • efa_schmid_leiman() changes:

    • For a lavaan solution whose general factor also loads directly on items, the general-factor loadings now include these direct loadings. Before, they were dropped with a warning, so an item that loads only on the general factor got a general-factor loading and a communality of 0. efa_reliability() and OMEGA() change too for such fits.
    • A lavaan solution with more than one second-order factor, other paths between the latent variables, or correlated disturbances now gives an error (efa_sl_higher_order, efa_sl_correlated_factors). Before, the transformation ignored these parts of the model, and the results were wrong without a warning.
    • The group factors now keep the column order and the names of the input solution, for example the factor names of a lavaan model, or "PA2", "PA3", "PA1" of a psych::fa() solution. Before, they were sorted by the number in their names and renamed "F1" to "Fk", so an indicator-to-factor map built in the input order could match the wrong factors. If the columns do not have unique names, they are still labelled "F1" to "Fk". SL() changes too.
    • For a lavaan solution, orig_R is now the correlation matrix that the model was fitted to, so efa_reliability() works with its default variance = "correlation".
    • g_name now defaults to NULL: the general factor is the latent variable that loads on the other latent variables. SL() keeps g_name = "g". A multiple-group lavaan fit now gives the error efa_sl_multigroup.
    • A standardized lavaan loading of 1 or more no longer gives an error. With correlated factors, such a loading can occur in a proper solution. A variance of 0 or less still gives an error.
    • With two first-order factors, the second-order loadings now get the same absolute value, and their product is the factor intercorrelation, as in psych::schmid(). No second-order EFA is run for them, and the warning tells this.
    • efa_schmid_leiman() no longer has .... The efa_fit() arguments that it passed on had no effect on the second-order EFA. Give the settings of this EFA in estimate_control().
    • A Phi given with a fitted solution no longer gives a warning (efa_sl_phi_specified). It replaces the fitted factor intercorrelations.
    • An efa_fit() solution with rotation = "bifactorQ" now gives the error efa_sl_bifactor, because it already has a general factor.
    • The warnings of the second-order EFA name the first-order factors also for a pattern matrix without names.
    • L2 is now a plain matrix with the column "g" and one row for each first-order factor, for every input. settings keeps only the estimator and the estimation settings of the second-order EFA.
    • print() shows the second-order loadings.

  • efa_simulate() with marginals = "empirical" changes:

    • Each dataset is now a new draw of independent cases: normal deviates at an intermediate correlation, each mapped through the empirical quantile function of its column of marginal_data (Cario & Nelson, 1997). The intermediate correlation of each pair is calculated once from the bivariate normal distribution. Before, the correlations of each dataset were fitted to the population correlation, so the datasets varied less than samples do. This made recovery in simulation studies look too good.
    • The data now reproduce the population correlation also with model error, or with an R that is not an exact factor structure. Before, only the common-factor part of the population was reproduced.
    • The rows are now in random order. Before, they were sorted by the last variable.
    • A population correlation that two marginals cannot attain now gives an error (efa_simulate_unattainable_cor). An intermediate correlation matrix that is not positive definite gives an error, or with force_pd = TRUE is smoothed, as for marginals = "VM".
    • n_factors is deprecated and ignored.
    • The results for a given seed are different.

  • efa_cd() generates its comparison-data populations faster, most of all with many variables. The populations have the same distribution as before, but the results for a given seed are different.

EFAtools 1.1.0

CRAN release: 2026-08-21

Comparing and Averaging Solutions

  • efa_compare() now saves print(), format(), and plot() display settings, and reuses them.

  • efa_compare()’s are_equal now returns NA, not 0, when the compared values differ before the decimal point.

  • efa_compare()’s elementwise differences are now signed.

  • efa_compare(reorder = "congruence") now also matches the sign of single-factor solutions.

  • Fixed efa_compare()’s handling of missing loadings. This previously inflated diff_corres and diff_corres_cross.

  • Fixed a bug that could trigger spurious warnings from efa_average()’s progress bar.

Data Screening and Simulation

  • Fixed two bugs in efa_screen()’s outlier detection. Both could give an incorrect or less robust result.

  • efa_screen()’s multicollinearity verdict is now based on the condition index alone, following Belsley (1991).

  • efa_screen() now withholds an uninformative Henze-Zirkler p-value instead of reporting an exact 0 or 1.

  • Mardia’s skewness and kurtosis in efa_screen() now use small-sample corrections.

  • efa_simulate() gains missing_vars, to restrict missingness to selected variables.

  • Fixed efa_simulate()’s reported population under marginals = "VM" and force_pd = TRUE.

Factor Retention

  • The type argument of efa_ekc() and ekc_type of efa_retain() no longer have any effect: the empirical Kaiser criterion now always follows Braeken and van Assen (2017). Suggested factor counts can differ from earlier results.

  • efa_nest() results can now be plotted.

  • efa_parallel() now shares simulated reference data between "PCA" and "SMC", roughly halving simulation time. Results for a given seed can differ from before.

Input Validation and Correlation Handling

  • efa_* arguments with fixed choices are now case-insensitive and accept unambiguous abbreviations.

  • Correlation matrices supplied as data frames are now recognised and analysed correctly.

  • Correlation-matrix smoothing now consistently returns a positive-definite matrix.

Missing Data and Multiple Imputation

  • efa_mi()’s rmsr_upper argument is deprecated and no longer used.

  • With target_method = "consensus", efa_mi()’s pooled result no longer depends on the order of data_list. Results can differ from earlier versions.

  • efa_mi()’s pooled RMSEA is now capped at 1, and its confidence interval is withheld when it excludes the point estimate.

  • An efa_mi() solution now reports the Heywood status of each imputation, in mi_admissibility.

  • A cor_method = "fiml" fit with analytic standard errors now withholds loading standard errors when the rotation is weakly determined, matching the other standard-error methods.

Model Estimation

  • The "uls" estimator now minimises the full residual, including the diagonal, matching its analytic gradient. Results can shift for over-factored models.

  • Starting values for the "paf", "ml", and "uls" estimators are now kept within [0, 1], matching psych::smc().

  • estimate_control() gains fiml_max_iter and fiml_tol, to control the FIML algorithm.

  • A cor_method = "fiml" fit now reports its convergence status in a new fiml component, and warns if it does not converge.

  • An RMSEA confidence interval that cannot be computed is now reported as NA, instead of stopping the fit.

Ordinal Correlations

  • Perfectly ordered polychoric or tetrachoric pairs are now reported as 0.9999/-0.9999, with a warning.

  • Binary pairs with an empty response cell now use a margin-preserving continuity correction.

  • Unavailable asymptotic variances for polychoric or tetrachoric pairs are now reported as NA, and affected robust standard errors are withheld.

  • Rare response patterns in strongly correlated pairs are now handled more accurately. They less often block DWLS or robust standard errors.

Power Analysis

  • efa_power() with group > 1 now always solves for a whole-number sample size per group.

  • In simulation mode, efa_power() now reports NA, not a misleading hit rate, when a retention criterion never gives a suggestion.

  • In simulation mode, efa_power() now rejects a p that disagrees with the population model, instead of silently replacing it.

Printed Output

  • Truncated variable names that would collide in loading tables are now abbreviated and numbered, so they stay distinguishable.

  • print()/format() for Schmid-Leiman loading matrices now honour max_name_length, name_style, sort_loadings, and max_factors_per_block.

Reliability and Factor Scores

  • efa_reliability() and OMEGA() now compute every omega total from the model’s full implied common variance, including cross-loadings and contributions from other group factors. Results can change for solutions with cross-loadings.

  • The whole-scale row of a correlated-factors solution is now labelled "total", not "g"/"general", which implied a general factor such solutions do not have.

  • efa_reliability() now recognises several solution types correctly, including Schmid-Leiman tables and single-factor solutions, which were previously misclassified, mislabeled, or refused.

  • efa_scores() and FACTOR_SCORES() with method = "Bartlett" or "Anderson" now stop on a communality at or above 1, instead of returning unusable weights.

  • efa_scores() and FACTOR_SCORES() now reject scoring data with a constant, infinite, or too-short variable, instead of silently returning NaN scores.

Rotation

  • Oblique rotations, including efa_procrustes(), now refuse a nearly singular transformation matrix at every step. Results can differ for near-degenerate solutions.

  • A gradient-projection rotation is now reported as converged only when its projected gradient meets tolerance. Convergence flags can differ from before.

Standard Errors

  • efa_fit(se = "sandwich") now withholds standard errors and confidence intervals when a Heywood case occurs.

  • Bootstrap output now reports the number of usable replicates when fewer than requested are available.

  • Analytic standard-error output now includes factor and variable names in SE$Phi and vcov_unrot_loadings.

EFAtools 1.0.0

CRAN release: 2026-07-23

New Interface

  • The recommended interface now consists of lowercase efa_* functions, with estimation and rotation settings defined through estimate_control() and rotate_control().

  • The existing uppercase functions remain available without warnings, so current scripts continue to work, but new features will generally be added only to the efa_* interface.

  • Estimators, rotations, retention criteria, scoring methods, and presets can now be specified without matching capitalization exactly, while results continue to store and display the standard spelling.

Changes When Using the efa_* Interface

  • The efa_* functions now report an error for unused, misspelled, or misplaced arguments instead of silently ignoring them.

  • Estimation and rotation options such as convergence settings, iteration limits, and rotation parameters should now be supplied through estimate_control() or rotate_control().

  • Factor-retention functions always fit unrotated models and therefore reject rotation settings that would not affect the retention result.

  • The estimator argument is now named estimator rather than method, and rotation arguments use the names p_type and random_starts rather than P_type and randomStarts.

  • efa_retain() requires the full argument name max_iter_CD, because the shorter form max_iter is no longer matched automatically.

  • The second-order estimation preset for efa_schmid_leiman() is now specified with estimate_control(type = ...).

  • Control objects are validated when they are created, allowing invalid settings to be detected before a model is fitted.

Reproducibility

  • A supplied seed now controls the complete analysis, including random rotation starts, bootstrap samples, group-specific fits, and model averaging where applicable.

  • efa_average() now has a seed argument and produces the same result regardless of the number of parallel workers.

  • Seeded results from efa_parallel() and efa_hull() no longer depend on the parallel-processing setup.

  • Because random-number handling and several calculations were corrected, the same seed may produce different results than in earlier versions for efa_parallel(), efa_hull(), efa_average(), efa_group(), and some efa_simulate() settings.

Standard Errors and Fit Statistics

  • efa_fit(se = "information") now returns NA with an efa_se_unreliable warning when standard errors for unrotated loadings cannot be estimated reliably, such as for Heywood cases or poorly distinguished factors.

  • Information-based standard errors are now calculated for correlation matrices rather than covariance matrices, improving their accuracy and agreement with sandwich and bootstrap estimates.

  • Information-based standard errors are now supported only with Pearson correlations and FIML; use se = "sandwich" for polychoric, tetrachoric, or rank correlations.

  • FIML-based analytic standard errors now use the same small-sample scaling as the package’s other standard errors and fit statistics.

  • efa_mi() no longer reports AIC, BIC, or ECVI when these measures are not meaningful for the underlying fitted models.

Multiple Imputation and Group Comparisons

  • The pooled unrotated solution from efa_mi() no longer depends on the order of the imputed datasets.

  • Pooled unrotated loadings from efa_mi() now use the same orientation conventions as a regular efa_fit() result, making the two easier to compare.

  • These changes can affect unrotated loadings, their standard errors, and statistics derived from them, but rotated solutions and their fit indices remain unchanged.

  • Consensus results from efa_group() no longer depend on the order in which groups are supplied, so congruences, loading differences, salience classifications, and invariance conclusions are now stable across group orderings.

Factor Retention and Rotation

  • Fixed a regression in the revised MAP criterion, which could previously suggest an incorrect number of factors.

  • efa_smt() now rejects polychoric and tetrachoric correlations because its tests and fit measures are not valid for these inputs.

  • Rotated factors are now labelled F1, F2, and so forth according to the variance they explain, rather than inheriting labels that could change with the random seed.

  • Rotation diagnostics now distinguish between the total number of starting solutions and the number that were fully optimized; code using n_starts should use n_starts_total or n_optimized instead.

  • Criterion-based rotations now respect the specified maxit limit during every stage of the optimization.

Model Averaging

  • efa_average() no longer prints NaN% when no models converge and now clearly states which fitted models are included in each reported rate.

  • Named presets such as "SPSS", "psych", and "EFAtools" now use their intended iteration limits, so models that exceed those limits are marked as non-converged and excluded from the average.

Data Simulation

  • Cudeck–Browne model-error simulations are now reproducible across computing platforms and numerical libraries, although a given seed will produce different populations than in earlier versions.

  • With marginals = "VM", categorized data now reproduce the requested category proportions more accurately, with a warning when the requested thresholds cannot be transformed safely.

  • The documentation now clarifies that match = "thresholds" and match = "polychoric" produce the same data for normal marginals, and the selected value is now stored in the result settings.

  • The documentation now explains the missing-data mechanism produced by missing = "MAR" and the residual bias it may create for analyses based only on the observed data.

  • The documentation now explains that model_error = "WB" generally produces a larger RMSEA for the generating model than the requested target.

Input Validation and Correlation Handling

  • efa_screen() now handles collinear or nearly collinear variables without failing during its outlier checks.

  • Polychoric covariance calculations now handle sparse or nearly perfectly associated response tables more consistently, allowing DWLS analyses to continue while warning about problematic variable pairs.

  • Covariance matrices supplied where raw data or a correlation matrix is expected are now rejected with a message suggesting cov2cor().

  • Polychoric and tetrachoric correlations now require ordered factors or numeric response codes, preventing character or unordered-factor levels from being interpreted in alphabetical order.

Power Analysis and Factor Scores

  • efa_power() now clearly documents that N is the total sample size across groups and returns the corresponding sample size per group in N_per_group.

  • Simulation-based power summaries now distinguish structure-recovery rates from the underlying Tucker congruence values.

  • The documentation for FACTOR_SCORES() now correctly describes R2 as squared factor-score determinacy.

EFAtools 0.8.0

CRAN release: 2026-07-07

New Functions

  • efa_group() performs multi-group EFA.
  • efa_power() to conduct power analysis (analytic power based on RMSEA and simulation based power analysis).
  • efa_reliability() to calculate various reliability indices.
  • efa_screen() to screen data for multivariate normality and suitability for factor analysis.
  • efa_simulate() to simulate data with various distributions and missing data mechanisms.

Changes to Functions

  • EFA():
    • gains cor_method = "fiml" for raw data with missing values.
    • gains se = "sandwich" and se = "information".
    • renamed the top-level fields boot.SE, boot.CI, and boot.arrays on EFA objects to SE, CI, and replicates, as the old names are no longer accurate given the additional SE implementations.
    • All criterion-based rotations are now computed by a built-in gradient-projection rotation engine instead of the GPArotation package. The rotated solutions are numerically equivalent but implemented in C++ for speed to allow high numbers of random starts.
    • The default number of random starts for the rotation in EFA() (randomStarts) has been raised from 10 to 100, making local minima less likely for the rotation criteria that are prone to them.
    • additionally reports the Tucker-Lewis index (TLI, also called the non-normed fit index), the expected cross-validation index (ECVI), and the standardized root mean square residual (SRMR)
    • gains a seed argument and now fits the non-parametric bootstrap replicates (se = "np-boot") in parallel across replicates via the future framework.
    • now has a summary() method that prints the full, detailed solution (loadings with communalities and uniquenesses, explained variances, fit indices, and residuals); print() now gives a more compact overview.
  • EFA_AVERAGE() gains cor_method = "fiml" (passed to EFA()).
  • EFA_POOLED():
    • now dispatches its multiple-imputation standard-error pooling automatically by the standard-error method of its component fits: Rubin’s rules for the information method (with Wald confidence intervals), the two-stage pooled-inputs (MI2S) approach for the sandwich method, and the existing bootstrap pooling for the non-parametric bootstrap method.
    • Renamed the top-level field boot.MI to MI.
    • now defaults to target_method = "first_target", which aligns every imputation to the first imputation’s rotated solution with a single Procrustes rotation. For orthogonal rotations this reaches the same pooled estimate as "consensus" with substantially less work. "consensus" (Generalized Procrustes Analysis of the imputation-specific rotated loadings) is still available as an opt-in for orthogonal rotations, and now raises a classed condition (efa_consensus_oblique_unsupported) when combined with an oblique rotation, because the iterated-oblique-Procrustes centroid can drift to degenerate targets.
  • cor_method now accepts "poly" and "tetra" to compute polychoric and tetrachoric correlations from raw ordinal (respectively binary) data, using a two-step estimator with no empty-cell continuity correction. Supported by EFA() (including its non-parametric bootstrap), EFA_AVERAGE(), the suitability tests KMO() and BARTLETT(), and the retention criteria EKC(), KGC(), MAP(), SCREE(), SMT(), and N_FACTORS(). The criteria that compare the data against simulated continuous reference data (CD(), PARALLEL(), NEST(), and HULL()) do not support "poly" / "tetra" and signal an error.
  • The factor-retention functions CD(), EKC(), HULL(), KGC(), MAP(), NEST(), PARALLEL(), SCREE(), and SMT() now return objects of a common efa_retention class with shared print() and plot() methods.
  • COMPARE() objects now have a plot() method that returns a ggplot object. The plot is no longer drawn as a side effect of print().
  • Console output (the print, summary, and format methods) is now produced with the cli package, and the messages, warnings, and errors emitted across the package are now S3-classed conditions, which makes them easier to handle programmatically.
  • EFA(), SL(), and EFA_AVERAGE() now accept method = "MINRES" as a synonym for method = "ULS". Minimum residual and unweighted least squares are two names for the same estimator and return identical results.

Bug Fixes

  • EFA() and EFA_POOLED(): The comparative fit index (CFI) now floors the model and baseline noncentralities at zero before taking their ratio (Bentler, 1990), so it is no longer deflated toward zero when the baseline (independence) model fits comparatively well. The value is unchanged for well-fitting models and now matches lavaan. CFI can change for solutions in which the baseline model does not misfit much.
  • EFA_POOLED(): Corrected and extended the pooled chi-square-based model-fit indices. The D2 average relative increase in variance is now centred on the mean of the square-root statistics (Li, Meng, Raghunathan & Rubin, 1991), removing a one-sided inflation of the pooled chi-square, RMSEA, and CFI. Bootstrap/MI confidence intervals for the pooled fit indices are now the Rubin-Wald multiple-imputation summaries; a miscalibrated bootstrap-percentile interval (obtained by re-running the pooling algorithm over matched replicate indices) was removed, as was a mislabeled pooled Fm.
  • EFA_AVERAGE(): When every averaged solution fails (all runs error, fail to converge, or are Heywood cases), the function now returns an empty (NA) averaged result instead of erroring or averaging an empty set.
  • EFA() with method = "PAF": the reported number of iterations (iter) is now the number of iterations actually performed; it was previously one too high. The loadings, communalities, and convergence status are unchanged.
  • ULS extraction: the minimised objective is now the sum of squared off-diagonal residuals, consistent with its analytic gradient and the reported minimum (Fm). The diagonal residuals were previously included in the objective but not in its gradient. The fitted loadings are unchanged to within optimiser tolerance.
  • NEST() and PARALLEL(): A failed eigendecomposition of a degenerate simulated matrix now raises a clear error instead of resulting in undefined behaviour.
  • The chi-square model-fit statistic is now the Bartlett-corrected maximum likelihood discrepancy evaluated at the model-implied correlation matrix, for both ML and ULS extraction. For method = "ML" this matches stats::factanal() and psych::fa(); the small-sample Bartlett correction was previously omitted. For method = "ULS" it is now a proper chi-square-distributed statistic matching psych::fa(fm = "minres"); previously the least-squares residual sum of squares was multiplied by N - 1 and read as if it were Wishart-distributed, which produced an invalid p-value. The independence (baseline) model used for the CFI is rescaled onto the same discrepancy scale. Consequently the p-value, CFI, RMSEA (and its confidence interval), AIC, and BIC change for ML and ULS solutions, and the number of factors suggested by SMT() and HULL() may change for these methods.
  • PARALLEL(): The percentile reference eigenvalues are now computed with stats::quantile(), correcting a slight off-by-one in the previous indexing.
  • EFA(): For oblique rotations, the factor intercorrelations (Phi), the structure matrix, and the explained variances are now reflected and reordered consistently with the rotated loadings. Previously, when a factor was reflected to a positive orientation the factor intercorrelations were not sign-adjusted (so the structure matrix and reported correlations did not match the loadings), and with order_type = "ss_factors" the factor intercorrelations were not reordered at all.
  • EFA(): The returned rotation matrix (rotmat) is now reflected and reordered consistently with the rotated loadings, for both orthogonal and oblique rotations, so that the rotated loadings can be reconstructed from the unrotated loadings and rotmat. Previously the sign reflection was not applied to it and its factors were left in a different order, so this reconstruction did not hold whenever a factor was reflected or reordered.
  • HULL(): The convex-hull elimination now tests every triplet of adjacent solutions, including the one formed by the last interior solution. Previously this final triplet was skipped, so a solution lying below the line connecting its neighbours could remain on the hull. This can change the number of factors suggested by HULL() (and hence by N_FACTORS()).
  • NEST(): When the test accepts every eigenvalue it examines (no empirical eigenvalue falls at or below its reference within the search range), it now retains that last accepted model rather than the model with one fewer factor. The search range is also bounded so that the reference model fitted at each step stays over-identified.
  • PARALLEL(): When every real eigenvalue exceeds its reference, so the decision rule finds no crossing, the suggested number of factors is now all retained components, reported with a warning, instead of a silent NA (matching the convention used by EKC()).
  • EFA(se = "np-boot"): The non-parametric bootstrap no longer repeats per-replicate warnings (about arguments pinned alongside type, and about the iterative fit reaching its maximum number of iterations) once per replicate. They are now suppressed during resampling, and non-convergence across replicates is reported once as a summary.
  • COMPARE(): With reorder = "congruence", the columns of y are now matched to those of x by an optimal one-to-one assignment that maximizes the total Tucker congruence, rather than by an independent per-factor best match. The greedy match could assign two factors of x to the same column of y, duplicating that column and dropping another, which corrupted the reported differences, factor correspondences, and root mean squared distance. The result is unchanged whenever the greedy match was already one-to-one.

EFAtools 0.7.1

CRAN release: 2026-05-08

Changes to Functions

  • EFA(): Added randomStarts argument passed to GPArotation functions, as suggested by Coen Bernaards.
  • FACTOR_SCORES(): Added rho argument, thanks to Andreas Soteriades.
  • EFA_POOLED(): Fixed issue that could lead to averaged Phi not being symmetric.

EFAtools 0.7.0

CRAN release: 2026-04-30

New Functions

  • MAP() computes the Velicer MAP criterion (both TR2 and TR4).
  • PROCRUSTES() to perform orthogonal and oblique Procrustes / target rotation.
  • CONSENSUS_PROCRUSTES() to perform Procrustes on a list of targets to obtain a common target.
  • residuals.EFA() to extract and print residuals and, if computed, standardized residuals.
  • EFA_POOLED() to run EFA on a list of multiple imputated datasets and pool the results. Thanks to Andreas Soteriades for the suggestion and first implementation.
  • print.EFA_POOLED(), print method adapted from print.EFA()

Changes to Functions

  • N_FACTORS() can also compute the MAP criterion.
  • EFA():
    • Now returns and prints residuals and, if SEs are computed, standardized residuals.
    • Calculates and prints RMSR.
    • Can calculate bootstrap standard errors and CIs of parameters and fit indices.
  • print.EFA() now prints more information.

Bug Fixes

  • Fixed a bug in OMEGA() that led to incorrect omega, H, and ECV values for lavaan bifactor models. Tnanks to Christopher D. King for bug report and suggested fix.
  • Small fix in the documentation of EFA_AVERAGE().
  • Fixed incorrect calculation of RMSEA in EFA().

EFAtools 0.6.1

CRAN release: 2025-07-30

Changes to Functions

  • EKC(): Now correctly returns the number of factors based on the first time the eigenvalues drop below the references values.

EFAtools 0.6.0

CRAN release: 2025-06-19

New Functions

  • Added NEST() to perform the Next Eigenvalue Sufficiency Test (Achim, 2017).

Changes to Functions

  • EKC(): The implementation based on Auerswald and Moshagen (2019) used in previous versions differed from the original implementation by Braeken and van Assen (2017). Now both versions are implemented and can be selected with the new type argument. Thanks to Luis Eduardo Garrido for pointing this out and to Johan Braeken for sharing sample code, based on which the original version is now implemented.
  • N_FACTORS():
    • Updated default settings to only use often used and well performing factor retention methods (others, like the Kaiser Guttman criterion can still be used).
    • Added NEST as additional factor retention method.
    • New arguments: ekc_type, alpha_nest, and n_datasets_nest to account for the changes in EKC() and to control NEST().

EFAtools 0.5.0

CRAN release: 2025-05-23

Changes to Functions

  • print.EFA():
    • Now returns explained variance from rotated, rather than unrotated solution, if a rotation was performed.
    • Now prints communalities and uniquenesses in loading/pattern matrix
  • EFA(): Calculate and return model implied correlation matrix.

EFAtools 0.4.6

CRAN release: 2025-03-21

General

  • Small change in test of .gof(): Changed some tests to take care of the ATLAS issue when using R-devel on x86_64 Fedora 34 Linux with alternative BLAS/LAPACK.

EFAtools 0.4.5

CRAN release: 2024-12-22

Bug Fixes

EFAtools 0.4.4

CRAN release: 2023-01-06

Changes to Functions

  • print.EFA(): Added arguments cutoff, digits and max_name_length that are passed to print.LOADINGS().
  • print.LOADINGS(): New Argument max_name_length to control the maximum length of the displayed variable names (names longer than this will be cut on the right side). Previously, this was fixed to 10 (which is now the default).

Misc

  • Updated a test of a helper function (.gof()) that threw an error when using R-devel on x86_64 Fedora 36 Linux with alternative BLAS/LAPACK.
  • added dontrun to examples of EFA_AVERAGE() and its print and plot methods as these were causing issues on R-oldrel which were not directly related to EFAtools and thus could not be fixed from within the package.

EFAtools 0.4.3

CRAN release: 2022-10-02

Changes to Functions

  • .gof(): Changed the helper function to take care of the MKL issue when using R-devel on x86_64 Fedora 34 Linux with alternative BLAS/LAPACK.

EFAtools 0.4.2

CRAN release: 2022-09-27

Changes to Functions

  • .is_cormat(): Changed the helper function to better detect wheter a matrix is a correlation matrix.
  • PARALLEL(): Added a check, testing whether N > n_vars and throw an error if this is not the case.

Bug Fixes

  • Fixed some tests due to upcoming changes in the psych package which EFAtools depends on.

EFAtools 0.4.1

CRAN release: 2022-04-24

Bug Fixes

  • Minor fixes in tests to solve problems on macOS m1.

EFAtools 0.4.0

CRAN release: 2022-03-21

Changes to Functions

  • EFA(): Changed error to warning when model is underidentified. This allows the Schmid-Leiman transformation to be performed on a two-factor solution.
  • OMEGA(): Added calculation of additional indices of interpretive relevance (H index, explained common variance [ECV], and percent of uncontaminated correlations [PUC]). This is optional and can be avoided by setting add_ind = FALSE.

Bug Fixes

  • CD(): Added na.omit() to remove missing values from raw data to avoid an error in the comparison-data procedure.

EFAtools 0.3.1

CRAN release: 2021-03-27

General

  • When testing for whether a matrix is singular and thus smoothing should be done, test against .Machine$double.eps^.6 instead of 0, as suggested by Florian Scharf.

Changes to Functions

  • EFA():
    • Added warnings if type = "SPSS" was used with method = "ML" or method = "ULS", or with a rotation other than none, varimax or promax.
    • Avoided smoothing of non-positive definite correlation matrices if type = "SPSS" is used.
    • Use Moore-Penrose Pseudo Inverse in computation of SMCs if type = "psych" is used, by calling psych::smc().
    • Use varimax_type = "kaiser" if type = "EFAtools" is used with varimax or promax.

Bug Fixes

  • EFA_AVERAGE():
    • Added future.seed = TRUE to call to future.apply::future_lapply() to prevent warnings.
    • Fixed test for Heywood cases from testing whether a communality or loading is greater than .998, to only test whether communalities exceed 1 + .Machine$double.eps
  • print.EFA(): Fixed test for Heywood cases from testing whether a communality or loading is greater than .998, to only test whether communalities exceed 1 + .Machine$double.eps
  • OMEGA(): Small bugfix when lavaan second-order model is given as input

EFAtools 0.3.0

CRAN release: 2020-11-04

General

  • Added examples for EFA_AVERAGE() to readme and the EFAtools vignette
  • Updated examples in readme and vignettes according to the updated OMEGA function

New Functions

  • Added function EFA_AVERAGE() and respective print and plot methods, to allow running many EFAs across different implementations to obtain an average solution and test the stability of the results.

Changes to Functions

  • EFA(): Defaults that were previously set to NULL are now mostly set to NA. This was necessary for EFA_AVERAGE() to work correctly.
  • PARALLEL(): Rewrote the generation of random data based eigenvalues to be more stable when SMCs are used.
  • OMEGA(): Changed expected input for argument factor_corres from vector to matrix. Can now be a logical matrix or a numeric matrix with 0’s and 1’s of the same dimensions as the matrix of group factor loadings. This is more flexible and allows for cross-loadings.

EFAtools 0.2.0

CRAN release: 2020-09-17

General

  • Created new vignette Replicate_SPSS_psych to show replication of original psych and SPSS EFA solutions with EFAtools.

New Functions

  • Added function FACTOR_SCORES() to calculate factor scores from a solution from EFA(). This is just a wrapper for the psych::factor.scores function.
  • Added function SCREE() that does a scree plot. Also added respective print and plot methods.

Changes to Functions

  • CD(): Added check for whether entered data is a tibble, and if so, convert to vanilla data.frame to avoid breaking the procedure.
  • EFA():
    • Updated the EFAtools type in PAF and Promax.
    • Added p value for chi square value in output (calculated for ML and ULS fitting methods).
    • Updated the SPSS varimax implementation to fit SPSS results more closely.
    • Created an argument “varimax_type” that is set according to the specified type, but that can also be specified individually. With type R psych and EFAtools, the stats::varimax is called by default (varimax_type = "svd"), with type SPSS, the reproduced SPSS varimax implementation is used (varimax_type = "kaiser").
    • Renamed the kaiser argument (controls if a Kaiser normalization is done or not) into normalize to avoid confusion with the varimax_type argument specifications.
  • ML(): Changed default start method to “psych”.
  • N_FACTORS():
    • Added option to do a scree plot if “SCREE” is included in the criteria argument.
    • Added a progress bar.
  • OMEGA(): Now also works with a lavaan second-order solution as input. In this case, it does a Schmid-Leiman transformation based on the first- and second-order loadings first and computes omegas based on this Schmid-Leiman solution.
  • SL(): Now also works with a lavaan second-order solution as input (first- and second-order loadings taken directly from lavaan output).

Bug Fixes

  • .get_compare_matrix(): Fixed a bug that occurred when names of data were longer than n_char
  • COMPARE(): Fixed a bug that occurred when using reorder = "names".
  • EFA(): RMSEA is now set to 1 if it is > 1.
  • HULL(): Fixed a bug that occurred when no factors are located on the HULL
  • KMO(): Fixed a bug that the inverse of the correlation matrix was not taken anew after smoothing was necessary.
  • PARALLEL():
    • Fixed a bug that occurred when using decision_rule = "percentile"
    • Relocated error messages that were not evaluated if no data were entered (and should be)
  • print.COMPARE(): Fixed a bug that occurred when using print_diff = FALSE in COMPARE().
  • print.KMO(): Fixed a bug that printed two statements instead of one, when the KMO value was < .6.

Minor Changes

  • OMEGA() and SL(): Added an error message if the entered term in g_name is invalid (i.e., it cannnot be found among the factor names of the entered lavaan solution).

EFAtools 0.1.1

CRAN release: 2020-07-13

Minor Changes

  • Added an error message in PARALLEL() if no solution has been found after 25 tries.

Bug Fixes

  • Updated different tests

  • Deleted no longer used packages from Imports and Suggests in DESCRIPTION

  • PARALLEL(): fixed a bug in indexing if method "EFA" was used.

EFAtools 0.1.0

CRAN release: 2020-07-07

  • Added a NEWS.md file to track changes to the package.
  • Initial CRAN submission