Changed the default value for n.rotations in fa from 1 to 20 to solve an interesting example data set from Keith Widaman. Added a discussion of multiple rotations in the help file.
Modified faRotation to distinguish between varimax (from stats) and Varimax (for GPArotations).
added vech function to misc functions
Modified the Rd file for rmssd to properly reflect the code. (reported by Maximillian Hagspiel)
Fixed a simple but nasty bug in describeBy when analyzing a single variable (reported by Brenden Schuetze – and many others)
Added scores as an option to irt.CFA to find irt based confirmatory scores.
Added irt.CFA to allow IRT based on the confirmatory factor model.
Added bootstrap resampling to CFA and CFA.bifactor to provide confidence intervals.
Replaced biplot.psych with biPlot. Deprecated biplot.psych
Added a trivial but useful function r2alpha
Added the ability to do fa.lookup on CFA output
Added a maxlength parameter to fa.lookup
Cleaned up documentation for psych
fixed a call to response.frequencies in alpha that should have been responseFrequency to avoid the deprecation warning. Reported by Jacob Junger
Fixed the call to print_psych.cfa to include the factor score correlation estimates.
Fixed a very subtle bug io describeBy when called by with – the bug was in char2numeric which worked on arrays, not single vectors. (Bug reported by Steven Zhou.)
Some Major changes include:
Added CFA to do confirmatory factor analysis. (Following the Spearman/ Guttman algorithm discussed by Dhaene and Rosseel, 2024). Uses keys or Lavaan syntax.
Added CFA.bifactor to extract a general factor and then do a CFA on the residuals following grouping information specified in the call.
Added a function (omegaStats) to report omega statistics for CFA.factor output. Generalized to the case of lavaan cfa output (where appropriate.)
Added a function (lavPars) to translate lavaan models into matrix form for various functions.
Added a new function (fsi) to estimate factor score indeterminancy following recent papers by Nicewander (2020)
Updated fa.congruence, anova.psych and diagram to handle CFA output.
Changed the default procedure for finding factor score indeterminancy from the Grice (2001) method to the Nicewander (2020) method. Kept Grice as an option in fa.
Added two additional outputs to cta.15 to allow it to plot histograms of tendencies to show that although the histograms appear normal, the bivariate plots do not reflect a stochastic process.
Added the zero order correlations in the output of lmCor and thus in the print and summary statements.
Started to deprecate redundant names for functions (e.g, set.cor, setCor)
Modified the factor determinancy index to more properly reflect Grice and then subsequently added the Nicewander procedure.
Added error messages to scoreItems and scoreOverlap to reflect improper input
Added a tolerance to isCorrelation so that the diagponnal can be very close to 1
Added CFA as option to diagram and omega.diagram
Added lavParse to sim and sim.structure to allow simulations of lavaan based models. Can specify the phi matrix in the lavaan model
Modified fa.congruence to include lavaan solutions.
Fixed the calculation of chi square in factor stats to recognize the symmetry of the residual matrix. Prior values were off by a factor of 2. Note, this does not change the other tests.
Fixed fa.parallel for the case of just fa output to match the both option (reported by Hyun Seok Do )
Fixed principal and fa to properly weight covariance matrix in the case of weights being desired (was just weighting correlation matrices). The problem with principal was reported by "CHIS".
Improved documentation for factor.scores to clarify that although fa uses regression weights for scores, factor.score defaults to tenBerge weights. (Thanks to Colin Toth for pointing out the inconsistency in the documentation.)
Corrected formula for Guttman 4 in SplitHalf (reported by Wolf Vanpaemel)
Fixed a problem with scoreOverlap when finding scores
The psych package includes functions and data sets to do classic and modern psychometrics and to analyze personality and experimental psychological data sets. The functions are written to further research in the Personality-Motivation-Cognition Laboratory at Northwestern University. The psych package has been developed as a supplement to courses in research methods in psychology, personality research, and graduate level psychometric theory courses. The functions are a supplement to the text (in progress): An introduction to psychometric theory with applications in R.
Additional functions are added sporadically.
This news file reports changes that have been made as the package has been developed.
To report bugs, send email to mailto:[email protected] using bug.report.
Version 2.5.7 is the development release of the psych package. It is available as a source file for Macs or PCs in the repository at https://personality-project.org/r/. The released version on CRAN is 2.5.6 (Added June 23, 2025). The second digit reflects the year (i.e., 2025), the third set the month (i.e., 2.5.6 was released in June of 2026 the last two digits of development versions reflect either an minor change or the day of any modifications, e.g. 1.8.3.3 was the third attempt to get 1.8.3 released. 1.7.8 was released in August, 2017.
To install the development version, use the command: install.packages("psych", repos="https://personality-project.org/r", type="source").
Remember to restart R and library(psych) to make the new version active. It is also useful to check the date of the development package, as I will use the same version number but change the dates (e.g., packageDate("psych")). The date of the CRAN version reflects when it was submitted.
Suggestions are welcome, but the current list includes the following (and has included some for a long time, so lets be patient):
Add confirmatory clustering to ICLUST
Add the ability to create multiple groups in sim.irt functions
Add omega factor extension figure option (requested by Sylia Wilson). Partly done.
Add option to do subject density and item density plot to IRT plot.
Add the ability to scale radar plots of raw data from min to max, and add a scale to radar and spider plots (e.g. circular histograms)
Check the bias in bootstrap resampling using cor.ci
Add symmetric=TRUE to eigen calls, think about only.values=TRUE for some cases
Add an analysis of DIF to the irt functions (requested by David Condon)
Add some power functions.
Add CIs for means and skews as an option (for describe as well as describeBy) Requested by Aaron Wichman – but I already report se so that ci for mean is trivial for user to find.
Add a function for the Meng, Rosenthal, Rubin tests for multiple comparisons of correlations.
Add statistic to fa to report max fa. (partly done with fm = minrank)
Parallize iterations for confidence intervals in omega
Probably should drop much of the stats when doing iterations with fa or omega
Find unbiased cis when doing bootstrap
Allow the choice between regression vs. correlation in esem between X and Y sets.
Improve documentation for esem
Add interbattery diagram to the structure diagram function.
Add Krippendorf's alpha? – probably not worth it
Add the ability to weight factor residuals to allow WLMS
Add the Hull method for number of factors. Probably not worth it in that it would give the solution too much importance. Number of factors is at best an estimate.
Add an option for the correlation type in bestScales (requested by Lorien Elleman).
Change help file examples so they are cleaner for people who example(fn)
Add testthat functionality - finally started 10/15/23 Making slow progress
Need to add documentation to mediate/moderate/lmCor for what dfs are used and why.
Fix bug in fa.lookup for omega g values that are negative (but when does this happen?)
Adjust stats in fa.extend to report the raw and then the extended values
Add just check variables specified for being numeric (instead of entire data.frame) (suggested by Fransisco Wilheim)
Pretty up the output from mediate (perhaps make a mediate2latex function)
Improve the print function for bassAckwards
Speed up fa.parallel for large number of variables (requested by Matt Zefferman)
Add VM correlation to lmCor
Bug reported by Juliius Pfaft in KMO. Actually, not a problem, but cleaned up code anyway. Improved the documentation as well.
Slight improvements to mediation.diagram Added error line to the dv and Added a = and b = to mediation.
Added true scores to the result list from sim.item
Added colnames to simulated data in sim and sim.item
Added cohenPooled to allow pooling of cohen.d/hedges.g across data sets
Added the ability to add longer variable names to fa.diagram using a dictionary.
Improved the documentation for fa.lookup to explain how to make a dictionary.
Added cohen.d to the objects that plot will plot (just calls error.bars)
Added plot.sum option to reliability to plot summary estimates by default
Improved the documentation for iclust.diagram to show how to label the variables.
Changed dia.self to plot errors from center of a dv rather than top. This solves the problem in lmCor that the error curves are misplaced.
2 errors in print statement for mediate (reported by Julian Basanovic)
The psych package includes functions and data sets to do classic and modern psychometrics and to analyze personality and experimental psychological data sets. The functions are written to further research in the Personality-Motivation-Cognition Laboratory at Northwestern University. The psych package has been developed as a supplement to courses in research methods in psychology, personality research, and graduate level psychometric theory courses. The functions are a supplement to the text (in progress): An introduction to psychometric theory with applications in R.
Additional functions are added sporadically.
This news file reports changes that have been made as the package has been developed.
To report bugs, send email to mailto:[email protected] using bug.report.
Version 2.5.4 is the development release of the psych package. It is available as a source file for Macs or PCs in the repository at https://personality-project.org/r/. The released version on CRAN is 2.5.3. The second digit reflects the year (i.e., 2025), the third set the month (i.e., 2.3.12 was released in December of 2023 the last two digits of development versions reflect either an minor change or the day of any modifications, e.g. 1.8.3.3 was the third attempt to get 1.8.3 released. 1.7.8 was released in August, 2017.
To install the development version, use the command: install.packages("psych", repos="https://personality-project.org/r", type="source").
Remember to restart R and library(psych) to make the new version active. It is also useful to check the date of the development package, as I will use the same version number but change the dates (e.g., packageDate("psych")). The date of the CRAN version reflects when it was submitted.
Added SRMR and RMSEA to the cortest functions.
line.col and ci.cor to pairs.panels following a suggestion from Jordan Adamson.
minor patch to describe to handle weird edge case of vectors of type character or factor (reported by Raffaele Mancuso)
Corrected the documentation and operation of biserial (reported by Ann-Kristin Koop). I had said x was dichotmous but actually it was looking for y as dichtomous. Now it will work either way.
As reported by Vlad Costin, the print statement in omega incorrectly reversed the stated probability values from 90
Fixed the print statement for undim to reflect the published article and the documentation.
Fixed lmCor so it will work on covariance matrices as well as correlation matrices
Increased the tolerance criterion for symmetric matrices in isCorrelation and isCovariance. Was causing a problem in esem example.
Improvements to omega function to allow specification of rotation option. Playing around with the cluster rotation as default.
Added rotation option to reliability (passed to omega).
Added the ability to have small (noise) factors in sim.hierarchical.
Added ... to the call to iclust.diagram to allow more control for lwd in diagram (Requested indirectly by Steve Reise.)
Modified the legend call in error.bars.by to allow flipping the order of the legend keys (thus allowing legends to be in the order (top to bottom) of the lines drawn.)
Changed the call to eigen in fa.parallel to just get the values following a suggestion from Niels Waller.
Improved bestScales to actually run in parallel. Also fixed the print function for bestScales to left justify content.
Added the hare function to rank choice voting
Fixed iclust so when called from reliability it will not choke on new.plot not called
Fixed error.bars.by to properly allow colors in cats eyes with multiple IVs
fixed reliability call to unidim to pass n.factors
Improved iclust.diagram to add control to figures.
At long last. added the ability to get cluster scores in ICLUST – analogous to factor scores (requested years ago by Ben Shalet). See help page for an example. ICLUST now returns the keys.list to be use for scoring.
Finally added Left justification to dictionary content in lookupItems and in bestScales
Added the ability to show probability stars in corPlot.
Added a short function (r2p) to convert r to p values.
Added the pval option to cor2
Added a somewhat more useful error message when polychoric fails. It now suggests setting correct=FALSE.
Added complexity to omega solution
Added an example of simulating a bifactor model to the sim help page.
Added RV to find the RV correlation between two matrices. Works with raw data as well as correlation/covariance matrices.
Added the min option to iclust when finding beta.
Added the ability to call reliability directly from ICLUST to get several alternative estimates of the general factor saturation.
Added the ability to process data without column names to reliability and splitHalf.
Improved the documentation to iclust discussing original versus purified scores.
Modfified iclust.diagram to have the ability to draw just alpha or just beta.
A number of changes to ICLUST including various patches to make the summaries work.
Added the goodness of fit statistics from fa.stats to iclust output.
Added Iclust beta values to reliability function.
Dropped the link to lme and xyplot in the help menu for mlr because this was causing problems for CRAN.
Made GPArotation depends rather than suggested to get around the no-suggests check on CRAN.
Strange bug introduced into score.multiple.choice by qgraph reported by Michael Truong. Fixed by changing class of describe object used in score.multiple.choice to just be a data.frame.
Fixed df2latex for short names not labeling h2 (reported by Alex Weiss )
Fixed a problem in polychoric which was leading to NaN values in rare cases due to a precision problem in cumsum of frequencies. Partly due to the correction for continuity. (Reported by Esmerala Ruiz Pujada with a nice data set that led to the problem.)
Fixed a minor problem in print_psych.bestScales for dictionaries with mislabeled columns.
Fixed alpha to properly report std.alpha in case of a keys vector (reported by Benjamin Bui)
Fixed problem in irt.fa that was not passing correct to the polychoric function
Corrected formula for BIC in the fa help page (reported by Jannick Orzek)
Fixed lmCor to properly calculate unweighted R between sets. (Was not correct for negative correlations)
Various patches to omega.sem related functions to allow them to work with improved omeaga
Fixed a bug in ICLUST where the final summary correlation matrix was wrong. Does not affect the cluster results.
Added use to the parameters of iclust. Had left this off in the 2.4.6.26 release.
Change examples to dontrun rather than commented out ## mostly done
Add the ability to get factor scores for higher order factors (requested by Brandon Vaughn). Partly done in omega, but there are problems associated with the entire problem of factor scores.
Fix a bug in alpha when finding average R (need to use cor rather than cov2cor when treating missing data)
Add weight option to statsBy (but the question is how to handle missing data)
Adjust the se for correlations (in bestScales) to take into account the pairwise count, rather than the n.obs
Improve documentation for thurstone. Would help if we had an example of scaling.
scoreIrt.2pl needs to be able to adjust for group differences correctly
Possible problem in pooled correlations for statsBy
Think about improvements in itemLookup to show means if desired
Fix bug in polychoric where it doesn't handle empty cells very well. (reported by Björn Büdenbender). Or at least improve the documentation about why this is a problem.
Fix bug in mediate for multiple dependent variables (reported by Martin Zeschke)
scoreItem should find the mean item correlation with na.rm=TRUE (Liz D)
create a scoreFastBy
improve the help menu for sim.multi and return the latent values
think about linking person scores to item difficulty scores in IRT
Consider using the Jennrich approach to standard errors of rotated loadings
Still need to fix mediate to handle more than 1 DVs
Possible problem with the use of try in several functions.
Need to fix omega so that alpha is correctly calculated in case of no general factor
allow omega style hierarchical structure for x (and y) in esem
Modified fa.sort to include cor2 output
Added the score option to scoreOverlap so that if given raw data, it will find the scale scores as well as the corrected for overlap statistics. Note that the scale scores are not corrected for overlap.
Minor tweaks to scoreIrt.2pl to check for bad data (missing cells ). This was leading to the function hanging forever in the case of pairwise counts of 0.
Added fix.dlpyr (as a local function) to statsBy and describeBy to not choke on Tibbles. Had added this to alpha several years ago.
Added itemSort to combine item means and item content and then sort them.
Added the median option to describe with the fast=TRUE option (requested by David Condon)
Parallelized describe. Now will describe 644K subjects with 6633 variables in 64 seconds (with fast=TRUE option). Improvement in speed varies by numbers of cores.
Parallelized bigCor for very significant savings.
Added the hist.border option to pairs.panels following a request by Jordan Adamson.
Added the w.exp (weighting exponent to cohen.kappa following a request from Julius Pfadt)
Added n.obs to unidim to properly find statistics from correlation matrices
Added a number of fit statistics to unidim
Added the ability to find Spearman correlations to bigCor. Added a check that the data are a data.frame rather than a data.table for bigCor.
Changed the class of describe output to be c("psych", "Pdescribe" ,"data.frame") to avoid choking if Hmisc is in use
Added the ability to use mean.weights in statsBy (still under testing)
Added paSelect to allow parallel analyses of subsets of items using a keys list.
Added vssSelect to allow VSS analyses of subsets of items using a keys list.
changed the number of elements in the unit.Circle in spider/radar to get more precision
Added a conversion to numeric in mediate to get around a problem with variables that are factors. (Problem reported by Lynn Bueeler )
Finally fixed bug in alpha where it finds R from cov2cor instead of directly (reported by Karl Ove Hufthammer )
Fixed item.lookup so that it will report item means as well as factor loadings.
Started adding test.that calls to various functions (requested by Coen Bernards) Slow but useful progress.
Seriously improved lavaan.diagram (and diagram) to be able to draw SEM diagrams correctly.
Modify lavaan.diagram so that it handles cfa output better (done)
The sim function was modified to allow for items as an option. This matches the parameters in sim.structure.
Changed the splitHalf function to do all comparisons, not just 1/2 (This was leading to a problem with small and odd nvar.) Discovered by Eunseong Cho.
Added the ability to specify type of correlation to cor2 and lowerCor
Renamed corr.test to corTest
Modified fa.extend so that it reports factor scores in a more reasonable manner.
Added part correlation (aka semi.partial) as an option to partial.r.
Added a discrete option to alpha to not complain about lots of categories.
Finally added the X and Y loadings for the canonical correlations in lmCor.
Added cancorDiagram to draw these canonical correlations.
Added canonical coeffiecients to lmCor (done 12/12/23)
Find canonical loadings in lmCor - done 12/12/23
Fixed reliability to find omega_h for 3 variables by defaulting to 1 factor
Fixed sim.structure so that it would treat low and high option correctly again. (Reported by Kubra Atalay Kabasakal). This option seems to have been dropped by accident in 2021.
The call in diagram to call lavaan.diagram now works (reported by Zhang Yaohua)
Fixed predict.psych for the lm case to allow correct prediction of unstandardized data
Added the n.obs parameter to corPlot to allow showing stars for correlation matrices
Added the ability to label higher order factors
added the coarse=TRUE option to factor.stats to solve obscure problem in simulating sapa data.
Fixed error.bars.by to allow drawing of v.labels for both by.var=TRUE and FALSE
Revised and improved unidim and the associated documentation. Unidim provides three estimates of the unidimensionality of a scale.
Fixed a bug in fa.diagram so that it will properly label the extension variables. This was actually a problem in dia.arrow for the draw=TRUE option
Fixed a bug (reported by Nadine Kaul and Maria Bobrowski) in corPlot where my ability to have slanted x lables messed up the xlas=2 paramter.
Also fixed a bug in corTest that was giving stars above the diagonal for the case of just lower diagonal output (also reported by Nadine Kaul and Maria Bobrowski )
Fixed bug in fa.multi.diagram to report variable labels correctly. (reported by Rémi Thériault)
Fixed a very obscure bug in describe where it would not return variable labels for vectors if asking for quantiles.
Fixed violin/violinBy to draw allow rain drops for ungrouped data
Added a threshold parameter to sim.hierarchical and sim.item to allow for conversion to binary data.
Added a n.small parameter to sim.minor to limit the number of small factors.
Added mean imputation to omega (for scoring scales)
Replaced solve(r,f ) with Pinv(r)
Added type option to mlPLot to allow for more control
Changed setCor.diagram to lmDiagram and added the ability to plot lm output.
Minor tweak to extension.diagram to allow the causal arrows to come from rather than to the extension variables.
Added rain as an option to violin to show rain cloud plots
Add the ability to slant labels in plots
Added a sort option to corPlot (sorts by cluster order)
Added RMSEA function to allow easy calculation of RMSEA given chisq, dof, n.obs (requested by requested by Gabe Orona)
Added RMSEA, but not yet TFI, etc. to esem (equested by Gabe Orona several years ago.
Cleaned up the various sim functions so that they are better documented.
Allow for different rotation options in X and Y sets in esem
Added the ability to return the true scores for subjects when simulating structures. This will help fitting reliability models This is now done for sim and sim.minor, sim.hierarchical, others are in process.
Changed a number of local functions in the form print.x to print_x format to comply with overly aggressive R development checks.
Fixed a very rare case in describeBy for the case of two categorical variables measuring exactly the same thing. Added char2numeric before the by(describe,group) call.
Fixed a bug in bestScales for the case of no items for replications (still in progress).
Fixed lmCor so that the degrees of freedom for the case of use="complete" is correct.
Dropped a redundant check in factor.scores for non-positive definite matrices. Pointed out by Chandler McClellan.
scoreIrt does not work unless keys are specified for nf > 1 (reported by Andrew Olney) – not sure of the problem. What ever it is, I think I fixed it.
Fixed a typo in the mlr rd file for the RkR equation. (reported by Sebastian Castro-Alvarez)
Modified alpha so that it returns a keys list rather than a vector
Added the smooth option to fa, fa.stats and fa.extension. This by default had been true. Now allows users to specify that they don't want to smooth improper correlation matrices. Affects just the fa.stats output.
Improved the documentation for irt.fa, and fa.extend
Changed the verbiage in omega to reflect that we are finding sums of squared loadings rather than eigenvalues (Thanks to Niels Waller for this point.)
Added the percentage of items with their highest correlation on the keyed scale (as an index of scale quality) Requested by Asghar Minaei.
Added the MIMS and MIMT matrices to scoreOverlap and to scoreItems These matrices reflect the average item correlation within and between scales (MIMS) and the average item x scale correlation within and between scales (MIMT). Suggested by Asghar Minaei.
Added options to the correlations in cortest and also improved the documentation.
Added alpha2r to convert alpha to average r (a trivial, one liner, but convenient sometimes).
Added crossValidationBoot to do multiple derivations / cross validations
Modified factorScore so that it will give scores even in the case of missing data. Scores are found from the mean of the non-missing items on a factor. To just get scores for non-missing subjects, specify impute="none".
Added faReg /faRegression to make use of fa.extend to find the regressions with dependent variables in a more understandable output. Modified extend.diagram to draw these as well. Although developed to use fa.extend, faReg will also do regressions based upon factor scores from the raw data.
Added SAPAfy to "sapafy" data matrices. (a trivial but useful helper function).
Improved examples for iclust.
Following a suggestion from Ethan Fosse, I have reversed the direction of the arrow in ICLUST to go from the items to the clusters. This is now the default but the old style can still be drawn using iclust.diagram with the bottomup parameter set to FALSE.
Minor tweaks to corPlot allowing the fixing of trailing zeros and the number of digits to display.
Improved corPlotUpperLowerCi to find correlations (and Cis) if given raw data. This improves usability, because the naive user would expect that behavior.
Added the show option to lowerCor so that it will just find but not display the correlations. Defaults to TRUE.
Added corCi and corr.test output to the objects that error.dots can display
Added eigenCi to find and plot the confidence intervals of eigen values.
Added more timings and changed the default size for bigCor
Added lmCor as an alias to setCor. (lmCor is much more intuitive as it is lm for correlation matrices).
Fixed statsBy to handle the case of NULL groups. (particularly for the cors=TRUE option). I think I have done this.
Fixed scoreItems so that the delete=TRUE option actually works. (Based upon a report by Christopher Wiese ). Also improved the documentation about why this is needed.
Fixed a typo in fa with the fm=alpha method. Reported by Philipp Doebler.
Finally fixed a bug in bestItems so that it works with raw data.
Fixed a very minor bug in polychoric that was throwing a warning and now is breaking. Results do not change. (Reported by Goran Kardum)
Fixed cor.smooth so that if it is given raw data, it will first find the correlations.
Fixed minor typo in fa.Rd (thanks to Ian Buller for pointing this out).
Fixed crossValidation to work with subsets of variables.
Fixed autoR to handle case of single variable when grouping. Reported by Alex Smith and thanks to various suggestions on stack overflow.
Fixed scoreItems so that it will give names to unlabeled data and keys (was breaking GAabbreviate)
Fixed crossValidationBoot for case of multiple DVs
Fixed reliability for the case of single item scales
Minor improvements to cohen.d to allow easier use of dictionary and sorting
Fixed cor.smooth to handle covariances matrices (reported by Douglas Otto as a problem in fa)
Finally fixed bug in bestScales so it will handle correlation matrices
Fixed bestScales so it will produce a keys list ordered by correlation for correlation input.
Fixed a bug in cohen.kappa for weighted Kappa. Reported by Ju, Yu-Jeng.
Fix a bug in bigCor that was not working for polychoric correlations
changed the use of the abline(h=1) in fa.parallel to draw it for pc not for fa. Pointed out by George Harrison.
Modified alpha so that it returns a keys list rather than a vector
Added the smooth option to fa, fa.stats and fa.extension. This by default had been true. Now allows users to specify that they don't want to smooth improper correlation matrices. Affects just the fa.stats output.
Improved the documentation for irt.fa, and fa.extend
Change the verbiage in omega to reflect that we are finding sums of squared loadings rather than eigenvalues (Thanks to Niels Waller for this point.)
Added the percentage of items with their highest correlation on the keyed scale (as an index of scale quality) Requested by Asghar Minaei.
Added the MIMS and MIMT matrices to scoreOverlap and to scoreItems These matrices reflect the average item correlation within and between scales (MIMS) and the average item x scale correlation within and between scales (MIMT). Suggested by Asghar Minaei.
Added options to the correlations in cortest and also improved the documentation.
Added crossValidationBoot to do multiple derivations / cross validations
Modified factorScore so that it will give scores even in the case of missing data. Scores are found from the mean of the non-missing items on a factor. To just get scores for non-missing subjects, specify impute="none"
Added faReg /faRegression to make use of fa.extend to find the regressions with dependent variables in a more understandable output. Modified extend.diagram to draw these as well. Although developed to use fa.extend, faReg will also do regressions based upon factor scores from the raw data.
Added SAPAfy to "sapafy" data matrices. (a trivial but useful helper function)
Added minor factors as an option to sim.congeneric.
Improveds the documentation for sim.minor to reflect the MacCallum paper. Improved the code so that it screens for impossible models.
Added/improved documentation for iclust.
Added the order object to iclust to allow for sorting matrices by cluster structure.
Added a sqrt option to VSS.scree and fa.parallel output (thanks to Marco Del Giudice for the suggestion)
Added a coefficients object for setCor and mediate
Fixed scoreItems so that the delete=TRUE option actually works. (Based upon a report by Christopher Wiese ) Also improved the documentation about why this is needed.
Fixed a typo in fa with the fm=alpha method. Reported by Philipp Doebler.
Finally fixed a bug in bestItems so that it works with raw data.
Fixed a very minor bug in polychoric that was throwing a warning and now is breaking. Results do not change. (Reported by Goran Kardum)
Fixed cor.smooth so that if it is given raw data, it will first find the correlations.
Fixed minor typo in fa.Rd (thansks to Ian Buller for pointing this out).
Fixed crossValidation to work with subsets of variables.
Fixed autoR to handle case of single variable when grouping. Reported by Alex Smith and thanks to various suggestions on stack overflow.
Fixed scoreItems so that it will give names to unlabeled data and keys (was breaking GAabbreviate)
Added the valid object to the pca function (to match the help file.) Reported by Tomosumi Haitani)
Fixed violin so colnames are not necessary
reliability function returns alpha from split half rather than from omega
Added the validityItem and CD.validity functions
Minor fixes to sim.minor to supply names for theta, minor clarification to sim.minor documentation.
rebuilt using R CMD build psych –compact-vignettes=both to compact vignettes
minor correction to pairs.panels to work in 4.2.0 without warning
Fixed splitHalf so it will work even in case of n.obs < n.var
Fixed error.dots to properly handle dictionary for some cases (reported by Matt Bales)
Added ability to do N random restarts in rotations in fa and added the faRotations function to do this more easily
Added a hyperplane count statistic to the output of fa (following Niels Waller), especially if we do multiple rotations.
Added congruence (aka cohen.profile) to find profile similarities and general congrucences.
Added scoreBy to find scores for multiple subjects from statsBy (requested by Lizz Dworak)
Added variance accounted for each variable in setCor (requested by Pierre-Emmanueal Encinar)
Added a citation to Grieder (2021) for a comparison of fa to SPSS
Added a comment in fa help that oblimin as implemented is a quartimin solution with delta =0
Added formula input to partial.r
Added m2d to find Cohen.d from two means and 2 sigmas (and 2 ns if desired)
Added bigCor and pairwiseCountBig to speed up correlations for large data sets. This takes advantage of the multiple cores available on Macs. bigCor is also useful for very large data sets even without multiple cores.
Modified setCor to just zero center the independent variables when taking product terms. (Suggested by Keith Widaman)
Added the confidence intervals for hedges g in cohen.d
Changing the name of categorical variables to *variable leads to problems. Added the flag option to not change the names in char2numeric
Fixed sim.multi to generate more than two factors (reported by Lizz Dworak)
Fixed reliability so that it doesn't give hopeless split halfs – was incorrectly keying items.
Fixed p value in mardia (don't multiply by 2!). Reported by Malen Roberts
Fixed the estimate of VIF for raw and unstandardized data
Fixed reliability so that it will work from correlations as well as raw data
Fixed bestScales to work from correlation matrices
Fixed a bug in describe for single variables from data frames (reported by Mike Linacre)
Fixed bug in cohen d reported by David Kellen (using wrong group size calculation)
Fixed bug in error.bars (and error.bar.by) to match the colors of the lines with the colors of the errors (reported by Lizz Dworak)
Fixed bug in df of ci for alpha.ci (reported by Mijke Rhemtulla.)
Fixed (again) the correct p values for ICC (reported by Mijke Rhemtulla.)
Modified the call to splitHalf from reliability to allow for sampling of all possible splits in the case of n.items > 16. The version released to CRAN (2.1.6) would unfortunately not complain if the n.items was too large and would cheerfully run for ever.
Also modified the reliablility function so it would not choke in the case of scales of size 1.
Changed splitHalf to only sample splits if the number of possible splits exceeds 10,000. (That means, if there are more than 16 items in a scale.) This can be adjusted in splitHalf but not in the reliability function.
Modified error.dots when using dictionaries to pull the column with a name of c("Item","item","Content","content"). It had been limited to "Content". This way you can use a full dictionary and not worry about the extra columns.
Improved bestScales so it will work with a dictionary more easily
Two further tweaks to reliability: it can be used on one scale without specifying keys. (Suggested by Tom Olino) and reports the unidimensionality statistic (u)
Think about adding the test for number of factors function from xx
Added factor.scores to the fa.extension so that we can do biplots
Various tweaks to radar to allow for more complicated graphics.
Added the ability to plot the variance accounted in bassAckward.diagram
Improved the help page of ICC to explain why the Shrout and Fleiss estimates are the same as the McGraw and Wong estimates.
Minor modifications to reliability and split half to return scoring keys for the best and worst splits as lists, rather than vectors.
Fix fa.diagram so that factors are enclosed in ellipses (as they had been). Reported by Mark Greenwood.
Added the normal=FALSE option to corr.test to allow for more precise estimation of probability values for method="spearman" or method="kendall". (Suggested by James Conigrave)
Added the ability to sort means correctly in item.lookup.
Removed the char2numeric in cohen.d (added in version 2.0.12) in response to an error report from Frank Papenmeier. It seems to have been an unneccessary fix to cohen.d.
Change the d2t call in cohen.d to reflect differences in sample sizes (reported by Emil Kiergagard). Also am now passing the hedges.g value when finding t
Fixed a slight disagreement between cohen.d and scatterHist for reporting the cohen.d. cohen.d was correct.
Fixed bassAckward.diagram to draw boxes for components, ellipses for factors.
divide n/2 in case of equal cell sizes in d.ci #reported by Fred Oswald
Added addStars to allow significance printout to corr.test (requested by Uthpala Pinto). Now, automatically gives a "stars" object. This has probability "stars" for raw and adusted p values.
Minor tweak to lowerMat to allow it to print character fields (for stars objects)
Adjusted the way p values are displayed for mardia (requested by Jhonys de Araugjo) to show out to 10^-300.
Modified the mediate function to report tests of ab terms against each other (requested by Nancy Collins)
Added within group density circles to scatterhist. A nice way of showing how M.distances work. Also added Mahalanobis distance calculation to scatterHist.
Added matrix as class to partial.r
Added various overlap measures for cohen.d (d2OVL, etc.)
Added the reliability function to find alpha, both omegas, and split half for a set of scales at once.
Added plot.reliability to allow for nice graphical displays of these results
The p and p.adj objects in corr.test were reversed for non-symmetric matrices (reported by wangshun)
Fixed isCorrelation to handle output from partial.r (reported by Isaac Petersen)
added unclass(r) to fa for the same problem of partial.rs
fixed catseyes to not choke if no variance.
Fixed a problem introduced into fa.diagram where it was not drawing circles for factors (reported by Mark Greenwood. )
Fixed a subtle error in bassAckward.diagram where the numbers where not quite alligning with the arrows.
Finally stopped getting zero length arrow warnings in error.bars
Fixed the mediate function to correctly report df for the b term (following reports of the problem from Nancy Collins and Amanda Montoya)
Fixed the printing of non-symmetric correlation p-values for corr.test (reported by Bowen Li)
Fixed predict.psych for the case of missing=TRUE to find colSums of weights time scores instead of colMeans. (No change in correlations, just the scale sd)
Added char2numeric feature in lowerCor and cor2 so they will no longer choke when given character information (e.g. sex, education level in sapa data)
Added flag=TRUE to char2numeric so that it changes the colnames of those variables that have been converted to numeric. Set flag = FALSE to not change colnames
Minor improvements to error.bars.by
Followed by a major rewrite of error.bars.by to make the code cleaner.
Added the data option to cohen.d so that we can just do a subset of the variables
Added the ability to do multiple DVs for anova of setCor objects.
Changed the sim (and related functions) to return the underlying theta that generates the data. This allows for tests of adequacy of recovery of the model.
Added pairwiseSample to allow for faster counting of pairwise frequencies.
Added pairwiseZero to display those pairs of items with 0 (entries) (Requested by J.D. Haltigan)
Completely rewrote sim.hierarchical to produce the orthogonal factor scores that produced the data. (Following correspondence with Niels Waller addressing the estimation problem.)
Added the adjusted probability as an output object in corr.test. (Requested by Alex Weiss).
Added legend to matPlot and other bells and whistles.
Modified testRetest so that it automatically sort on time and user id and drop subjects with non-matching tests. (Inspired by some comments from Jacob Lee).
Reordered spi.keys to match the bassAckward order. Added Big and Little 27 identifiers to the spi.dictionary (in psychTools)
Added to the fa and smc documentation to reflect using pseudoinverse.
Fixed statsBy so the grouping variable can be categorical and don't use it when finding the within group correlations.
Improved testRetest to allow for a different number of subjects in the trial 1 and trial 2 data. Also allows for randomized data.
Fixed a bug in describeBy to handle more than 2 grouping variables (reported by Nicholas Stafniak)
Changed the term eigenvalue to sum of squared loadings in omega (in response to a comment by Niels Waller)
Fixed bug in testRetest (reported by Jacob Lee) about fixed id column
Released 2.0.9 to CRAN 10/05/20 – thus 2.0.12 is now the development version
Added the option to print correlations in lookupItemFromKeys (requested by Seo Young Myaeng)
Modified isCorrelation to allow for NAs in correlation matrices (requested by Lorien Elleman)
Improved error.bars and error.bars.by in response to requestes by Lizz Dworak
Improve violin and violinBy to provide more features including vertical versus horizontal plots, stripcharts of the data points, and error bars.
Added fa.pooled in response to a query from Asghar Minaei about how to pool various imputation runs.
make cohen.d handle categorical (character) data (requested by Evan Grandoit)
Finally got the s.e. to intercept in set cor (requested by Eric Youngstrom) to work. See attempts from version 1.9.12 onwards.
Added a anova.test for two setCor objects (did this a while ago, not sure when)
Fixed a bug in setCor where it was not correctly handling the case of std=FALSE, zero = TRUE (reported by Robert Stupinsky)
Fixed a bug in mediate.diagram so that it properly ignores the intercept (also reported by Robert Stupinsky )
fixed a hidden function in thurstone so that it actually works given ranked data
Fixed a bug in scoreFast for the case of one scale and count.responses=TRUE (reported by Alyssa Counsell)
Finally fixed a bug in error.bars.by for the case of multiple variables with some missing data
ptci in corCi should be 1-ptci
Corrected the spelling of Branscombe in the documentation of Garcia (with thanks to Barbara Toizer)
Added drop=FALSE to violinby so it can handle single x variables.
Replace all calls to mixed.cor with mixedCor (responding to a question from Lisandro D Colantonio about mixed.cor being deprecated in a call from statsBy)
Finally (?) fixed bassAckward.diagram to draw the lower level factors correctly both vertically and horizontally
Fixed a bug in faBy so it will work with one factor (reported by Kondwani Kajera)
Fixed print.psych.vss to properly handle NA in BIC and EBIC (as reported by Paulo Sergio Panse Silveira )
Finally ? got the error terms in setCor for the intercept and the residuals to match lm!
Seems to be something wrong in mediate df 1 IV (fixed?) )
Added formula input to densityBy
Added grouping option (and formula input) to scatterHist
Fixed a minor confusion between main and title in fa.plot
Add minlength parameter to lowerCor and lowerMat for length of abbreviations
Fixed a problem with crossValidation where it would not work with a single criterion
Fix a problem with alpha if the number of categories differed drastically and max was not specified. Reported by Daniel Bradford.
Fixed a bug in geometric.mean where it was not processing na.rm for the case of a single vector (reported by Jesse Aldersliesten)
Fixed a bug in mediate where the standard errors of the boot values were wrong (reported by Felicia Zerwas)
Added legends to densityBy
Modified unidim to give cleaner output. Added two new fit statistics as well.
Added cross references to fa.multi with fa and omega.
Dropped sem as a suggested package. Basically, I was no longer using it, and had replaced the functionality with lavaan. This makes psych a lighter package since sem indirectly requires a number of packages.
A problem in fa or fa.stats reported by Kyle Husman (something is wrong with promax solutions) was actually a problem in factor collapse in coarse coding. Modified fa.stats so it pads out NAs for those correlations that don't exist
Continued the addition of formula mode to various descriptive functions (describe, describeBy, densityBy)
Fixed a bug in iclust.diagram introduced in 2.0.7 (substitute does not work work on vectors)
Finally fixed a bug in testRetest where I was flipping two of the components of variance (id x time and ID x items) Reported by Miquel Biron=Lattes
Added a warning when doing omegaSem with one factor (amd made it work – reported by Erich Studerus)
Fixed the order of plotting variables in bassAckward.diagram (was plotting them in reverse order)
Fix scoreItems to think more carefully about what to do if an item has no variance. Score it anyway, just don't use it for item statistics. (raised by Eric Green). It turns out the delete=FALSE option is scoreItems was doing this anyway.
Found a fixed a bug in alpha where the imputation was not working quite right.
Improved summary and print functions for faBy (since I was working on it anyway)
Added simCor (alias for sim.correlation) which allows for several types of skewing of the data, as well as dichotomizing it at cut points.
Modified iclust.diagram and more importantly, multi.arrow to address a problem introduced in version 2.0.7 which although it worked in R.4.0, broke in R.3.6.3. The problem was that the change from 3.6 to 4.0 introduced a change in the way strings are treated as factors. This is set by default to be FALSE in 4.0, but needed to set stringsAsFactors=FALSE when forming data frames when doing multiple graphics. The implication of this is that 2.0.7.31 will work for 3.6 and the version on CRAN (2.0.7) works for R >= 4.0. Note that another problem was introduced in the iclust.diagram function by this fix. (see psych 2.0.8)
Fixed bug in faBy so that it realizes we have done a statsBy already (reported by Lisandro Colantonio )
Added crossValidation to be compatible with setCor and bestItems
Added the weights option to statsBy (requested by Joshua Tuttle) – still needs to be worked on for weights within groups
Minor fix to error.dots to allow more control on the figures and to make sure that we are calling psych::describe
Now reports the keys used in scoreItems
Added the spengler data set as an example for mediate
Improved faLookup to handle bassAckward output
Improved the print and summary from mediate to include the intercept term.
Improved bassAckward.diagram to allow labeling of higher level factors and to take advantage of the improvement in diagram.
Improved several of the diagram functions (e.g. bassAckward, fa.diagram, iclust.diagram etc.) to draw the variable boxes and the path labels as vector calls. This leads to an impressive increase in speed when drawing the figures.
Improved char2numeric to handle case of characters not stored as factors
Improved mediate to use multicores (if available) to speed up the bootstrapping for big problems.
Improved describe to handle case of characters not stored as factors
Added responseFrequency as alias to response.frequencies and added class psych.
Added the ability to use formula input for describe (calls describeBy) and describeBy (see the revised help page)
Added formula input to cohen.d and cohen.d.by (allowing a cleaner call to cohen.d for 1 grouping or 2 grouping variables)
Added formula input to violinBy, densityBy, histBy.
Added the ability to lookup which items in a dictionary have a certain content (lookupItems)
Added the num.cat option to polychoric for those determined users who want more than 8 categories.
Replaced solve with Pinv in the partial.r function. This allows for inverses of matrices that are not positive semi-definite.
Commented out several tests for bestScales in order to pass the Debian timing tests on CRAN
Fixed summary for setCor to properly print betas (coefficients) Error introduced in 1.9.12.31 (reported by Hans Marius Norbom)
Fixed bug in bassAckward for orthogonal rotations (reported by Eva Billen )
Fixed bug in fa.congruence for case of factor loadings provided by user rather than a psych function (reported by Lisa Debruine)
Changed the way to find the pooled correlation in statsBy by finding the average fisherZ transformed correlation (back transformed to r) rather than just using the average weighted r.
error.bars (which was calling error.bar.by) was not passing all the parameters to error.bars.by. (Reported by Lorien Elleman).
errorCircles was not plotting error circles if there were NAs in any of the descriptive statistics\
Fixed the equamax call to cfT to include kappa in the call (reported by David Condon and Gerard Saucier)
Fixed passing parameter to Equamax (reported by Erich Studerus)
Fixed alpha to correctly report the alpha if dropped values for the case of nvar=2. I was right for the first item, not the second (reported by Thomas Buhl)
Fixed bug in statsBy that was not returning ci.wg for the case of just two variables (reported by Kate R)
Fixed typo in spelling of confirmatory of omegaSem
Fixed a bug in bestScales for the case of very few items (reported by Tim Kaiser)
Finally fixed a bug in polychoric where it would not properly correct for continuity in the case of different number of response alternatives. (Partly following a suggestion by Ben Mainwaring. )
Changed the solve(R) to Pinv(R) in cohen.d when finding Mahalonobis D.
Modified the examples in bestScales to not test while doing CRAN installation, in order to pass the Debian timing constraints.
Changed smc to us Pinv rather than solve to find the inverse.
Changed isCorrelation to isCovariance in alpha to allow us to find alpha from covariance matrices
Changed isCorrelation to check for symmetry, diagonals all = 1 and values -1 <= x <= 1
Added the y option in polychoric to match tetrachoric and the various cor functions. This allows asymmetric correlation matrices.
Added a number of different correlation options to pca/principal to match those in the fa function.
Added setCor to the models that can be used for prediction in predict.psych
fixed tetrachoric to handle y variables as well as x variables. Further modification to correctly handle empty cells.
Fixed fa.lookup to properly handle negative signs in names of omega output
Fixed response.frequencies to handle tibbles following a suggestion from Mikko Ronkko.
Moved various larger data sets to psych tools
Added summary function for cohen.d
Added summary function for testRetest
Added col option to AUC to allow for black and grey AUC curves
Item improved irt.stats.like an added make.irt.stats to allow importing external item information (Requested by Zara Wright)
A considerable improvement in speed in pairwiseCount by using crossprod instead of normal matrix multiplication
Added the ability to find confidence intervals for the r.wg and r.bg in statsBy (requested by Sandrine Muller)
Added the ability to find confidence intervals from a correlation matrix using the corCi function.
Added a note to corPlot about the problem with correlations > 1 (pointed out by Ryne Sherman). Then added the option to make all values > 1 the same color.
Added a new function (scoreWtd) to allow for weighted scoring. Also improved bestScales to find weighted as well as unit weighted scores.
Added a diagnostic message and error counts in multilevel.reliabilty for the alphaBy function following a query by Randy Lee.
Added sim.bonds to simulate the Thomson (1916) bonds model of independent factors with cross loadings.
Finally changed the default for corPlot to be numbers=TRUE.
Modified setCor and mediate to find the intercepts and standard errors of the intercept.
Modified scoreWtd to allow for input from lm or setCor
Modified error.dots and cohen.d to allow for error.dot plots of cohen.d results
Added a sort option for cohen.d
Improved bestScales so that it can report scales (both weighted and unitweighted) for multiple values of n.items (from low to high). Suggested by Lorien Elleman and Sarah McDougald.
Added aslist as an option to factor2cluster
Added the intercept term in mediate
Implemented a small change in fa2latex to format h2 as h^2 (requested by Alex Weisss)
Added the ability to compare models using the anova function for setCor and mediate.
Added the ability to specify item difficulties and discriminations for scoreIrt (requested by David Condon and Zara Wright)
Problem with scoreIrt.1pl with keys of length 2
Confidence intervals in ICC were incorrectly based upon alpha/2 instead of alpha. (reported by Wei Chen). Fixed.
added main to plot call in mlr
added sqrt of communalities in fisher.test in circ.tests (reported by Lisa Barrett and Jiahe Zhang)
Take the square root of communalities in polar
Fixed scrub (introduced a bug in 1.9.4 that prevented it from working)
Fixed bestScales to handle matrix input (detected by Anne Zola)
Minor tweak to statsBy so it will not find r.bg correlations for two groups (since this is meaningless).
Corrected the definition of S/N in the writeup for alpha. (Error reported by Carla Leal Kaymalyz)
Finally fixed mediate to handle 2 or more DVs
Fixed a bug in factor.residuals to pass Phi to factor.model. (Reported by Asghar Minaei)
Painfully changed from using class(x)[2] to inherits(x,value) to respond to changes in R 4.0.0 in most functions.
Added the Phi to the fa object in the bassAckward output to allow for factor output with loadings, correlations, and SS loadings. (In response to a question by Niclas Kuper.)
Added dictionary capabilities to manhattan
Added top as an option to dummy code. This will give just the dummy codes for the top most frequencies
Added a check for keysList and items in scoreIrt.2PL and others for the case where we mistakenly reverse them
Added the ability to take correlation input for manhattan
A serious improvement to bestItems so that it actually is useful
Added lookupFromKeys as a useful helper function
Added TargetT to do orthogonal target rotations. (It was already there, just not visible.)
Added faRotate to allow many types of rotations (also already there, just not visible). In response to a suggestion by Alex Weiss.
Added Target as an option to faCor (so that we can compare target rotations)
Added correction = .5 to AUC in response to a problem with 0 cell entries reported by Jeff Crukley.
Added the 'mollycoddle' feature to scoreItems, scoreFast, scoreVeryFast and setCor to check for bad input. (Suggested by Tim Bates).
Added guttman.lambda2 to splitHalf (requested by Muirne Paap)
Added Pinv to find pseudo inverses to allow calculations in schmid (however, this is not actually necessary). (Adapted from ginv from MASS.)
Add scores to the output of bestScales in units of criteria (actually done in the predict function) (requested by Lorien Elleman)
Added the ability to use "schmid" as a rotation option in bassAckward (requested by Joanna Hartung )
Added v.lab option in error.bars.by (as requested by Maike Luhmann )
Modified scrub so that it will process categorical data.
Added a bestScales prediction to the predict.psych function following suggestions by Sarah McDougald.
Cleaned up bestScales so the code is a bit cleaner. Improved the documentation for bestScales and predict.psych to show how prediction works with bestScales.
Added the vars option to biplot so that we can draw arrows to the data or to the variables.
checked the BDI scores in affect – they seem strange, but are in fact correct. They are mean scores.
check the eigen values in iclust - something is wrong
Fixed problem in Hedges g (reported by Chuck Powell)
Fixed a problem in isCorrelation for a square, symmetric matrix which is not a correlation matrix. However, we needed to fix this again by using all.equal.
Fixed a problem with mlr/multilevel reliability for the lme decomposition. Variances were indexed by location rather than by name. This led to a problem for certain cases. (Reported by Chelsea Dickens)
Fixed manhattan so that it will plot -log p values that are inf (by replacing with the largest non-infinite -log p
Fixed inconvenience (not really a bug, just poor documentation) in alpha with complete keys specified by location instead of name (reported by Richard Pond)
Fixed a bug in corr.p where it was not labeling the confidence intervals correctly. (Reported by Jeffrey Lees.)
Fixed bassAckwards for problem with orthogonal rotations (reported by Joanna Hartung))
Fixed bug in RMSEA calculation in factor.stats introduced in 2017 (reported by Michael Lee Crowe )
Fixed bug in statsBy for reporting correlations with multigroups. It was not labelling the results correctly. (reported by Eva Skimma)
There seems to be a problem with fa.parallel when using multi-cores on Microsoft Open R. See the note for fa.parallel for a fix. )
Added cohen.d.expected to find the expected value of the M.dist. It is sensitive to sample size and thus has a non-zero expected value.
Added dictionary option to cohen.d to include a list of items used.
Added the ability to do MIMIC diagrams from lavaan output in lavaan.diagram
Added the ability to specify the size in ellipses drawn by scatter.hist (requested by Jacques-Daniel Basic )
Modified dfOrder to allow a global direction (ascending, descending) option, and also to handle matrices
Added the ability to use stats in error.dots (comparable to error.bars)
Continued to improve bestScales
Minor changes to lookup and selectFromKeylist to make compatible with bestScale issues
Added the log.p option to bestScale
Invisibly return the correlation matrix (if found) from corPlot
Added matSort as an alias to mat.sort
Changed the documentation in cohen.kappa following a suggestion by Kelly Reeve
Added the fparse (formula parse) function to the namespace to make it public
Modified setCor and mediate to handle quadratic terms if specified in the formula input.
Added manhattan to do Manhattan plots of correlations versus a criterion.
Added the ability to impute data in alpha
Minor changes in fa.organize to allow for matrix input
Minor improvements to esem and esem.diagram to allow for lavaan model output.
Added bassAckward to do the bass-ackward approach of Goldberg (see also Waller). This returns the correlations of factors across solutions and draws a diagram.
Added faCor to find the correlation and congruences between two factor solutions of the same data.
Added extension.diagram to make cleaner fa.extension diagrams. Can be called from diagram()
Improved the documentation for fa.extension to explain more applications.
Added EPI personality scales, as well as film and drug condition to msqR. Re-arranged the order of variables in msqR to be more user friendly.
Changed the message in fa.stats about bad solutions from a message to a warning (which can be suppressed using suppressWarnings).
Fixed omega.diagram when drawing from a loadings matrix
Fix ylim parameter in plot.irt so that it actually works.
Fixed setCor so that it can handle single x variable with raw data.
Fixed corr.test so that it actually does use min.length
Fixed alpha so that it properly warns when it needs to reverse keys (incorrectly omitted this feature in the 1.8.8 release. Reported by Mohammad Latifi)
More importantly, fixed alpha so that it actually follows the help file. That is to say, it now will score short subscales appropriately.
Fixed bug in setCor when doing partial rs.
Fixed a related problem in mediate/moderate/setCor in terms of df when doing partial r.
Fixed a problem in mediate when doing just regressions with no mediation and just the correlation matrix.
Fixed biplot.psych so that it keeps the correct frame of reference so that abline(h=0) will work. (Reported by Fabio Daniel Trinco.)
Minor tweak to setCor to properly print unweighted multiple R
Investigated possible bug in cohen.kappa confidence intervals (reported by Marco Fornili). Output agrees with CohenKappa of DescTools nd ci.kappa of statpsych.
Changed the link from icar-project.com to icar-project.org to provide more stability
Improved the graphics of structure.diagram (tweaked the e.size parameter).
When the fix for fa to handle covariances correctly was made in 1.8.9,(September) another error was introduced that led to improper estimates for small problems. This has been corrected.
Fixed problems with generating code for lavaan and sem for correlated models in structure.diagram
fixed a bug in dia.arrow that was causing problem lavaan.diagram problem
Added minlength=5 to corr.test (requested by Felipe Dalla Lana)
Added robust version of cohen.d (d.robust)(See algina, 2015)
Modified keys2List so that it does not reorder the variables in the keys. This is useful when we use scoreItems and want to use the corrected.cors for estimating the correlation matrix, or other uses.
Improved bestScales so it specifically does k-fold cross validation and bootstrap aggregation (bagging). Improved the documentation a bit. That is, for bestScales consider the implications of .623 sampling due to bootstrap for the cross validation sample. Why do 90-10 when we are already doing 63.2/36.8?
Added a summary for bestScales
Added a few more descriptives to alpha (min.r and max r)
Added Niels Waller's direct Schmid Leiman solution (see Psychometrika 2017) as a rotation option and the function directSl.
Using Niels function, also added omegaDirect to calculate omega this way.
Added Procrustes rotation (from Niels Waller)
Added a few more options to fa.lookup
Fixed omega.diagram so that variables are sorted by F* factors.
Cleaned up omega.diagram to allow for omega.extension output and omegaDirect output.
Added intercept and confidence intervals for betas in setCor (requested by Franz Strich)
Added an nfactors option to fa.parallel (requested by Florian Scharf).
Added the add option to error.dots (allows for overplotting with dotcharts)
Added the cluster option (now the default) in fa.organize
Added the ability to plot single data vectors in error.dots
Added grouping variable to dummy.code.
Added scoreVeryFast to find sum scores for large data sets.
Parallelized scoreFast and scoreVeryFast for faster processing.
Added pairwiseImpute to impute correlation values for structured data
Added pairwiseReport to identify particular item pairs with low counts
Changed the name of count.pairwise to pairwiseCount to be consistent with family of pairwise functions (count, describe, impute, report)
Clarified the parameter settings in r.test (following a request by Cassandra Brandes)
Added the order variable parameter for error.dots to allow forcing a sorted order.
Modified sim.structural/sim.structure so that when simulating items, it more closely matches the factor model requested. It was generating the data and then simulating the items with one more pass through randomness, leading to structures that did not meet the requested structure.
Modified error.bars.by to take formula input and to draw two grouping variables appropriately
Added m2t to find t.test from 2 means and 2 standard deviations
Added cex and l.cex to mediate.diagram, moderate.diagram, and setCor diagram following a suggestion by Shu Fai CHEUNG
Modified multi.hist to give more control over margins and titles
Improved scoreOverlap to allow for raw data input (but, of course, this will still not result in scores).
Modified bestScales to name the best.keys object. Improved the documentation to emphasize the general utility of the function.
Fixed documentation in AUC which did not have specificity correct
Fixed Draw tetra to correctly find (phi) and to report the 2 x 2 matrix
Fixed alpha so that it names the keys when it returns them. (not a bug, just a pain that it did not).
Fixed bug in mediate diagram for 2 DVs, one IV
Fixed setCor.diagram to plot residuals for the correlations for the DVs
Added weights option in fa call to tetrachoric and polychoric (thanks to Suzanne Dufault for the report)
Fixed the bug in print.psych.mediate that was improperly reporting the indirect effect for the second IV.
Fixed a problem in fa where with the covar=TRUE option and covariances > 1, the results were incorrect (reported by Phillip Doebler).
Corrected code in cosinor to handle radians correctly (reported by Obioha Durunna)
Fixed fa.parallel so that the default for quant is .95 and is actually used.
Fixed principal so that the oblique target rotations are labelled as transformations (Ti instead of Ri) (Reported by Wu Hao)
Fixed a problem in scoreOverlap and scoreItems for the case of a single, negatively keyed, variable for a scale. Problem due to an infelicity of the diag function.
Fixed bug in splitHalf (not correctly reporting the mean split half. Reported by Wes Bonifay.) The problem was we were not counting the first split.
Fixed the dates of various parts of the news file so that they read correctly, and thus the Version > "1.8.4" trick works
Fixed the sign of t2r for negative ts (detected by Ashley Brown)
When using multiple raters in cohen.kappa, report the average (Light's kappa) as well as the individual pairwise kappas.
Added fromTo to allow choosing variable sequences by names of first and last
Added cs (taken from Hmisc:::CS function) to convert non-quoted strings into quoted vectors.
added acs to convert input into quoted strings.
Added median r to alpha, scoreItems, scoreOverlap, splitHalf (this serves as a signal for poor scales if the median.r differs very much from the av.r)
Improved bestScales to allow easier bagging of results.
Improved the documentation of bestScales to emphasize the bagging nature of the function. Separated the documentation from fa.lookup and lookup.
Fixed a problem with describe when describing a single column data frame. Reported by Vilmantas Gėgžna. The problem is with the way sd handles data.frames.
Fixed mediate so it will do regressions without mediation ala setCor.
Fixed a bug in corr.test for the case of single x or y variables (reported by Alex Weiss)
Fixed a bug in alpha where it was not working from a correlation matrix with n.obs specified. (Reported by Nicholas Stefaniak).
Fixed some problems in bestScales reported by Elizabeth Knowlton.
Minor changes in documentation (e.g. in alpha, violinBy, corr.test, ...)
digits option to print(x, all=TRUE) in print.psych
Modified violinBy to treat character variables as numeric (with notification) and to adjust the minimum and maximum in the case of groups. (Following some suggestions by Christie Nothelfer.)
Added the omit option to describe to just describe numeric data (another suggestion by Christie Nothelfe)
Modified p values in corr.test to allow for very small p by using the log.p option (suggested by Nicholas Clark)
Modified read.file and write.file (etc) to allow file= as well as f=. This is to make it more obviously compatible with save.file().
Dropped all sex > 2 values in spi.
Improved the dictionary for spi to include the demographic codes.
Added the density option to error.bars (suggested by Anne Zola).
Added the na.rm=TRUE option to dummy.code (suggested by Elizabeth Knowlton)
Added the ability to have criteria as a separate object in bestScales
Added formula option to setCor
Improved the input to setCor and mediate to allow for formula input. Also allowed interaction terms in setCor for parallelism with mediate.
Improved bestScales so that it summarizes the results of iterations. Fixed a problem with the overlap option so that it correctly chooses the number of items.
Improved dfOrder so that it sorts dataframes more readily.
Greatly improved setCor and mediate to allow for multiple interaction terms and to include higher level interactions. Both functions now return the transformed data when doing interaction terms. Improved the print.psych function for setCor.
Added a HowTo for mediation and moderation, cleaned up the other HowTos to allow them to cross reference each other.
Added the Garcia and Tal_Or data sets for demonstrations of moderation and mediation.
Added the ability to partial out variables in mediate.
Switched to using the log option in pnorm for probabilities in pt and pf
parallelize iterations when bootstrapping fa (did this some time earlier, but had not taken if off the to do list)
Added alpha option to corPlot (to allow semi-transparent plots)
Dropped the density =50 option and changed to alpha=1 in violinBy. This leads to much faster plotting.
Added the densityBy function to draw multiple histogram/density plots to show the overlap between two or more distributions. Can also just draw the density for one distribution.
Modified the call for bi.bars to be consistent with densityBy and violinby calls.
Minor improvement in vignette documentation
Added the AUC (area under the Curve) function. This basically adds signal detetection capability as an option for dichotmous data (and reports such measures as sensitivity and specificity as well as d').
Added alpha.ci to calculate confidence intervals for alpha based upon Feldt (1987)
Added confidence intervals to splitHalf function.
Added the df in mediate
Modified the call sequence to matReg to make it more transparent in mediatethe individual ab paths in the mediate function (requested by Heather Urry)
Added the describeFast function for quick counts of large data sets.
Added the filesList function to display all the files in a directory (using file.choose to find the directory).
Added the fileCreate function to do what file.choose() should do for RStudio
Added the filetype option to read.file and the ability to write without a suffix to write.file
Added filesInfo for information about the files in a directory
Completely rewrote the dfOrder function after finally understanding the help menu for order.
Added testRetest to do test-retest reliability and a number of related statistics.
Added checks for non-numeric data and data with no variances to mixedCor.
Added a check in mlr so that it will only find within subject alpha if n.times > 2 (This was throwing a warning if using mlr to check testRetest results.)
Serious rewriting of mixedCor with the hope of making it faster (and more robust to cases where we have unequal number of categories in polytomous data). Perhaps solved the problem of poly.di blowing up if the polytomous have a different number of categories.
Fixed the degrees of freedom in setCor and mediate to reflect the actual numbers of subjects in the correlations, not the number of cases in the file.
Modified ICC to use lmer from lme4 to handle missing data and to report variance components. Set the default to use lmer. This is particularly useful if doing large files.
Added the theta parameter to the sim.irt family of functions to allow for specification of the theta across simulations. (Requested by Yuming Liu.)
Added DAT as a possible data file type to read.file.
Cleaned up the diagram.Rd file demonstrations to look nicer.
Added a min.length parameter to corPlot to allow for abbreviations of row and column names.
Added the sai and tai data sets (state anxiety and trait anxiety measures from the PMC lab) for demonstrations of test-retest reliability, validity, etc.
Changed all the http:// references to https:// to get around problems with failure to read http:// files.
Added ci <- sef <- ci.adj <- NULL in corr.test to fix problem reported by Nicholas Clark 9/13/17
Fixed an inconsistency in mssd between the single vector case and the matrix/data.frame case. Was dividing by n-lag -1 instead of n- lag.
Clarified documentation for mssd (problem raised by Dennis Mongin)
Discovered that write.file(new=TRUE) works only in R.app and not in RStudio. This seems to be because core-R ignores the new=TRUE option.
Correct was not being passed correctly in fa for doing polychoric or tetrachoric correlations (detected by Jason French, 10/10/17)
Fixed improper error message bug in cohen.kappa reported by Johannes Bracher
Fixed spi.dictionary so that it labels variables correctly.
Fixed a print error in mediate for the a path (probabilities were displaying incorrectly.)
Made the digits option actually work in pairs.panels (reported by Federico Filipponi)
Fixed a bug in fa for some problems (e.g. fa(bfi[1:5], covar=TRUE)
Fixed a problem with diagram.curve and another in diagram.curved.arrow
Fixed a problem with plot.fa.parallel for the case of plotting just one line. (reported by Daniel Morillo).
Fixed the lavaan and sem models for omegaSem to handle reversed items properly (this involves gsub for - signs in names).
Fixed the ab boot values in mediate when we partial out variables (reported by Dan Molden)
Corrected the BDI scores in the affect data set. They had been incorrectly the EA1 scores.
Modified the help file for read.file to explain more clearly how to read .rda files.
Modified polychoric so that when it tests for number of categories, it is not limited by 1-5 or so, and can handle 0,20,40,80, 100 if these are few enough categories.
Changed the (1-alpha) "% confidence" to (1-alpha)*100 "% confidence " in error.bars.by (requested by Victor González Fernández)
Fixed the residuals reported in fa.extend (not giving correct values for uncorrelated factors –reported by Deon DuBruin) (and modified the print statement
Changed the degrees of freedom in paired.r to be n-3 to match r.test (and Steiger 1980)
Fixed setCor so that the ruw values are correct in the case of negative correlations. (reported by David Condon).
Added left and right brackets to variable names in df2latex to allow stars to print correctly in LaTex
Corrected the definition of the objective function in principal (error pointed out by Andre Beauducei)
Fixed cor2latex (and df2latex) so that large negative correlations can be bold faced as well as large positive correlations (pointed out by Anne Zola).
Fixed the label option in error.dots so that it actually works.
Completely rewrote partial.r for readability and also to add the ability to find the anti-image (that is to say, to partial all the variables from each other in correlation matrix.)
Added some options to autoR.
Added the gr option (for color gradients) to corPlot following a suggestion from Anne Zola and Lorien Elleman
Added fa.random to start doing random effects factor analysis.
Added cohen.d to find cohen.d statistic for two groups on a number of variables. Also reports the Mahalanobis distance.
Added Mahalanobis distance for groups (see del Giudice, MBR, 2017)
Added d.ci (cohen.d.ci) to find confidence intervals for effect sizes.
Added d2t, t2d to do basic conversions of effect size.
Rewrote bestItems and bestScales to allow raw data or correlation matrix input and to be much faster.
Added the pc option to fa.diagram to allow the arrows to come from the observed variables if the data are a principal components solution. (suggested by Michael Wood).
Added selectFromKeys to allow selecting items based upon a keys.list
Added fa.sapa to allow iterated factor solutions across random subsets of subjects. Basically just a tool to examine the stability of sapa based solutions.
Completely rewrote pairs.panels to a) add confidence regions for regressions and for loess fits. Changed the smooth function to be loess rather than lowess. Adjusted the histograms so that values line up with the labels. Changes suggested by Julian Martin.
Added alpha factor analysis as an option to the fa function.
Converted package to be byte-compiled. This seems to result in a 7 - 10 % improvement for e.g., factoring a 1000 x 100 categorical data matrix (mean of 61 elapsed vs. 65 secs).
added sem.diagram and sem.graph which call structure.diagram and structure graph with sem output. Still having problems sporadically with RGraphviz.
Added in cross validation as an option to best.scales/bestItems.
Modified alpha so that it can just select a small subset of items to score
Added a comment to the r.test function about precision of p values (suggested by Julia Rohrer).
Add an additional object to corr.test and corr.p to give Bonferonni and Holm (default) adjusted cis (requested by Lorien Elleman)
Added the name for corPlotUpperLowerCi to be the camel case equivalent of cor.plot.UpperLowerCi
Changed the se.bars=TRUE to se.bars=FALSE in fa.parallel to make the decision rule more transparent
Modified fa so that it will not die if nfactors = nvar
Changed messages to warnings in polychoric. This allows fa.parallel to run in parallel (and therefore much faster) without crashing. (Based upon a problem reported by Guido Biele).
Added the diag option to residuals and resid.
Added the blant data set to show factor analytic problems.
Fixed problem in cohen.kappa to correctly find the variances of weighted kappa (reported by Lisa Avery).
Fixed mssd to correctly divide by number of cases-lag instead of n.obs - l - lag (reported by Charlotte Vrijen)
Fixed problem with alpha for two item scales (I was not correctly reporting the alpha if one item was dropped.) Reported by Daniel Zingaro.
Patched cohen.kappa so that it warns instead of breaking if given identical ratings for everyone (reported by Caroline Wehner).
Added the factor names for old.min(res) solutions.
Fixed (finally?) a bug in mixedCor (coming from polydi) that was incorrectly reporting half of the correlations as 1.0. This was causing improper solutions in any data set that had more than 1 dichotomous variable and more than 1 polytomous variable.
Fixed scoreIRT.2pl so that it does not choke on 1 items scales
Modified fa so that it will report the factor score correlations for the chosen scoring method as well as the regression based scores. Although factor.scores was reporting the correct R2 values, fa was always reporting the regression based R2 values. Also deprecated the oblique.score option. (Based upon trying to answer a question from Jeannette Samuelsen.)
Added read.xport to the read.file so we can read SAS export files. Also added the widths parameter so that read.file can read a fixed width file.
Added select option to scoreItems and scoreOverlap to just score the items selected (to enhance speed for big sets of items and short scales).
Added a private function (selectFromKeyslist) to handle numeric and character keys.lists when doing selection.
Added irt.se to find standard errors based upon the test information
Added show = FALSE to setCor to allow the option of showing the unweighted matrix correlation (Had removed this in prior release, but decided for teaching purposes, is is convenient.)
Added the ability to define the separator when reading csv files in read.file
Added breaks=21 to multi.hist to allow for better plotting control
Added the ability to plot asymmetric correlation matrices in corPlot (requested by Anne Zola)
Added the correction option to fa.parallel when doing tetrachoric or polychoric correlations. It had been set to .5 without the ability to set it to 0. (Inspired by a question from Ismail Cukadar).
Added stars as an option to pairs.panels and corPlot. This is for those people who insist on showing 'significance' values. Also added scale as an option when showing the numbers (to show significance by number size.)
Added quickView to complement headTail and topBottom. Added from and to parameters to all three functions.
Added top=4 and bottom=4 to replace hlength and tlength in headTail and topBottom. Makes more obvious code.
Added Vaccounted for fa and pca (as requested by various stackoverflow users).
Dropped the message about matrices not being square in cortest. (Just an irritant when doing R CMD check and testing the intro.rnw file)#Adapted from the help for pairs #modified December 15, 2011 to add the rug option #further modified March 30, 2012 to add the method of correlation option (suggested by Carsten Dormann). #Fixed a bug in show.points on March 18, 2017 (reported by Matthew Labrum) #by moving all the little functions to be outside the main function, this allows method and rug to be passed to these lower order functions. #this should allow for somewhat cleaner code for the other functions #modified March 15, 2015 to add the ability to control the size of the correlation separately from the cex variable in the points #also added the ability to set the number of breaks in the histograms #Also completely reorganized the main function to be much cleaner #would like to add the smoothScatter function as an option
Modified plot.parallel to go from ymin rather than 0. Suggested by Hao Wu.
Added confidence intervals for the ICC1 and ICC2 to the output for statsBy (requested by Lorien Elleman)
Added mixedCor as an improvement to mixed.cor. Somewhat cleaner interface.
Modified the minres solution in fa following very helpful suggestions from Hao Wu. Basically, the first derivative wrong and the solution was slightly incorrect. This has now been fixed for minres. fm="old.min" keeps the old solution. The fm="uls" uses empirical first derivatives works to give the unweighted least squares solution. This is discussed in the fa help page as well.
Fixed RMSEA error (as reported by Hao Wu). In version 1.6.12 I had changed the RMSEA formula from the Tucker formula to the one used in sem and now have reverted to the Tucker formula (as described in the help page.)
Fixed cor.plot.upperLowerCi so that it works again and also reports replications
patch to mixed.cor for figuring out what kind of data we have. Basically, set all minima to 1 so that polydi won't croak
moderate.diagram was drawing incorrect X paths.
plot.irt was not labelling factor numbers
t2r was not taking the sqrt and thus was giving much to small rs. fixed.
In describeBy, changed nrow to NROW to correct error thrown by the affect data set for the development versin of linux (although that seems to have been a temporary error.)
Modified isCorrelation yet again to treat the case of symmetric correlation matrices stored as data frames. (Used in some class examples!)
Added sim.multi to simulate and draw within subject data.
Minor modifications to corPlot for labelling (and added a comment to the help file to show how to do multiple plots to answer a query from Yuji Shimohira Calvo)
Modified cohen.kappa so that the number of levels can be specified if there are some levels that are missing completely (requested by Amy Finnegan).
Added multilevel.reliability (mlr) to find various generalizability coefficients for three way (subjects x time x items) data following chapters by Pat Shrout and Sean Lane.
Added a helper function (mlArrange) to convert semi-wide data into long data. Useful for multilevel analyses of three way data sets.
Added mlPlot to plot multilevel (id x time * item) data.
Added a plot option to fa.parallel to allow suppression of plotting (requested by David Copndon)
Added autoR to find autocorrelations based upon mssd. This is useful for multiple time points.
Added a trivial helper function r2c (and cor2cov) to convert a correlation matrix to a covariance matrix.
Added VIF (Variance Inflation Factor) which is just 1/(1-smc) to setCor.
Added a note to alpha about the dplyr problem and added a fix to the problem (as reported by Adam Liter).
Added a discussion to the fa.rd documentation discussing the problem of estimating the minimum residual. (Inspired by correspondence with Hao Wu and Mikko Ronkko).
Added a new vignette (intro) and modified the overview vignette. Meant to make the introduction and overvies easier to read.
Added a new function: unidim to estimate the unidimensionality of a set of items.
Inspired by a question from Jared Smith, added several new features to scatter.hist, as well as removed a few bugs.
While patching pairs.panels (see bugs fixed), add smoother option to all the other options.
Added cor2cov (and then temporarily removed because of conflict with lavaan.shiny call to lavaan::cor2cov).
ScoreIrt: corrected cut so that it passes through to lower level functions (reported by David Condon)
rownames for corr.p$ci were incorrectly labeled (reported by Lorien Elleman)
Fixed a bug in finding RMSEA and the confidence intervals thereof. (Reported by Hao Wu). Possibly the error was introduced in 1.6.12
Fixed plotIrt so that it correctly rescales ylim for each factor rather than just the first factor.
The lag parameter in mssd was not being used. Fixed.
The handling of missing data by groups in statsBy is fixed for finding ICC1. This was producing seriously incorrect ICC1 and ICC2s with a great deal of missing data. That is for SAPA data, when grouping by ZCTA, many items were missing. (Reported by Lorien Elleman).
Fixed a bug in scoreItems and scoreOverlap when keys are already a data.frame rather than a matrix. (Reported by Jeromy Anglim).
Cleaned up various problems in mediate and print.psych.mediate so more complicate problems will work.
Modified isCorrelation to test for symmetric matices but dropping colnames and rownames. (This led to a problem with the neo data set. Reported by Hao Wu.)
Changed the neo data set from data.frame to matrix.
Fixed omegaFromSem and print.psych.omega to handle case of 1 factor sem model (which would normally be thought of as inappropriate, but some people wanted.)
Fixed a problem with printing describeBy output. (Reported by David Uttall.)
Fixed a problem with esem for the case of having more variables in the correlation matrix than being analyzed in the esem statement.
Fixed a bug in pairs.panels where it was ignoring the show.points option in several option cases. (Reported by Matthew Labrum).
allow scaling of circles in error.circles (suggested by Josh Wilt)
Added char2numeric to solve a problem of categorical data from a questionnaire.
Added a helper function dfOrder to sort data.frames on multiple columns.
Added interbattery for interbattery factor analysis.
Changed the print line for MLE chi square to say model based chi square.
Added Promax as a possible rotation in kaiser, and added the m = option to it as well. (In response to some correspondence with Stefan Dombroski).
Clarified the KMO and cortest.bartlett help pages.
Started to get rid of all of the deprecated functions. (or at least move to them to unique help files).
Added a trivial but useful function isCorrelation to improve the tests in alpha, scoreItems, etc.
Serious reworking of omegaSem and omegaFromSem to allow the use of lavaan or sem to do cfa as part of omega. Also, it is now possible to do omega directly from a sem/lavaan run.
Added lavaan as a source of omegaSem and omegaFromSem. Also, for omegaFromSem, added some of the normal omegasubscale output (suggested by André Kretzschmar).
fixed alpha, principal, fa, kmo and omega to check if the input is a symmetric correlation matrix or just a strange case of raw data (following a suggestion by Adam Liter).
Modified cohen.kappa to truncate confidence limits to +/- 1 and added a discussion to the help page describing how to do a boot strap of cohen.kappa to examine the distribution of cohen.kappa (suggested by Andreas Spitzmueller).
allowed scoreItems to handle variables that are factors. Added a warning when scoring total scores with impute="none" when missing data are found.
Added a parameter to densityBy (violinBy0 to allow for scaling the widths by sample size.
Added a helper function, dfOrder to sort (order) a data.frame by multiple columns
Completely revised scoreIrt (and the associated functions) to get around problems with all 0 or all 1s and to make the function faster, and perhaps more accurate.
Changed the sort option in irt.fa to be FALSE (makes scoring easier).
Fixed VSS so that it reports not the biggest, but also the first where the sign changes.
Allow imputation in factor scores (e.g. median or mean imputation for items) (requested by Stephan Daus). Also report the number of missing observations per subject when doing factor scores.
Fixed the RMSEA in fa.stats. Although the confidence intervals were correct, the RMSEA value was slightly inflated. Reported chi squares were correct, just the RMSEA was slightly off.
omega and schmid incorrectly estimated total variance in case of being given a factor loading matrix and correlation matrix as input instead of the more normal case of a raw data set or a correlation matrix.
Bug fixed in scoreIrt where the values for the scores for people who gave all of the highest possible scores was incorrect (reported by Roland Leeuwen ). In addition, the scoreIrt.poly was not properly doing normal scoring but was in fact just doing logistic scoring. Fixed. This bug was affecting those people with max or min responses, and thus was particularly a problem for short scales.
Fixed a bug in scoreItems for cases where some variables (not scored) are categorical (Reported by Ronald Fischer).
Fixed bug in fa.parallel where it is not drawing both resampled and simulated lines. (reported by Alexander Weiss). This was intentionally added in about March to make a cleaner diagram, but now I am discussing this in the help file.
Allow specification of the direction of items in the scoreIrt.2pl to match the keys.list (previously it would match the factor analysis which could be backwards for some scales with many reversed items).
Did not need to fix bug in wtd.table in polychoric by adding in minx and maxx, etc. (reported by Jeovani Schmitt) because this can be handled by settting global=FALSE.
Minor fix in the default labels for the fa.parallel function (suggested by Meik Michalke)
Made it clearer that score.irt with the keys option is basically a Rasch model.That is, tau and equal loadings are specified by keys.
Serious rewriting of scoreIrt (both for dichtonous and polytomous items) to make it able to use multiple cores (speeding up at least 8 times) and making the code more efficient when scoring multiple scales.
Added scoreIrt.1pl and scoreIrt.2pl to more easily find Rasch like and 2 parameter solutions.
Added scoreFast to (very) rapidly find mean or total scores with or without imputation (ala scoreItems) but not to bother finding all the complicated statistics which take a lot longer to find. In addition, it just bothers to look at the items that to be scored, rather than all the items in the items set.
Added a trivial function (cor2) to correlate two or more sets of data (x and y), round to 2 decimals, save and print the results. Truly trivial, but a time saver when comparing solutions.
Added more functionality to dotchart and created error.dots.
Added r2t,t2r,chi2r,r2chi as trivial conversion functions
Added read.file, write.file, read.file.csv, write.file.csv as helper functions for input/output
Added the ability to do Minimum Rank Factor Analysis (following a request by Michael Paul Grosz)
Added an option to print.psych to allow significant digits to be controlled for describe (requested by Jeremy Biesanz)
Added an exploratory structural equation function (esem) to do exploratory modeling, and esem.diagram to show it.
Added keys2list which converts a scoring key matrix to a list form.
Added the ability to just pass a keys.list to scoreItems, scoreOverlap, scoreIrt.1pl. This clearly simplifies the call.
Added another helper function: keys.lookup that takes a keys.list and a dictionary and displays the content of the items to be scored.
Did to cor2latex what we did in df2latex (i.e., add cut and big options)
Fixed vss to do any kind of rotation (that is, just pass the rotation to fa).
Modified error.crosses to allow specifying text, point, and error arrows separately (suggested and with help from Arnaud Defaye)
Minor change to tr to allow it consider na.rm=TRUE, and for cohen.kappa to handle cases of items with no variance.
Fixed subtle problem when handling missing data for score.irt for raw scores. Prior versions (< 1.6.7) would report incorrect total scores for cases with missing items that were to be scored negative.
Corrected the reference to Revelle, Wilt and Rosenthal in the bfi data set.
Fixed a subtle problem in print.psych.iclust that was causing errors in the development version of R for PCs.
Added mu backinto sim.hierarchical (thanks for Alan Robinson for noticing it was gone).
fixed bug in cor.wt for the problem of a vector of weights rather than a matrix of weights (Reported by Emil Kirkegaard)
Added an iterative option to cosinor so that it can do phase adjustment.
Added error.bars.tab to do plots of tabular data with error bars
Added a sort option to irt.fa so that items are sorted by factor loading.
Added pca as an alias to principal
Added the cut and big option to df2latex
Added a few cross references to make it easier to find functions.
Modified describe to take character vectors and matrices
Added labels to factor.plot for more than 2 factors. Not as hard as I thought. Added a show.names options to plot the labels without overwriting a point.
Modified corPlot to find correlations if input is a non square matrix or data.frame (but not an object from fa or omega).
Modified describe (and therefore, describeBy) so that it can take more options (including interquartile range) and quantiles (requested by Gilbert Gregory).
Added a weight option to fa to weight subjects by a weighting vector (requested by Mike Hammer).
Added anova.psych to allow for tests for chisquare differences between factor solutions.
Serious modifications to biplot.psych, fa.plot, and cluster.plot to allow names (labels) in the multi-panel (more than 2 factor) option. Also added the choose option to plot just one selected pair of factors from a larger set. This is useful if showing a 3 or 4 factor biplot and then "blowing up" one of the pairs.
Serious analysis of the use of global and correct for examples of small data sets provided by Gabriele Cantaluppi. Some resulting improvement in documentation and options.
Changed principal (and fa) so that when it calls GPA::rotation(Varimax) it passes options (such as eps=1e-7) to get more accuracy. (Reported by Gottfried Helms)
Fixed error.bars.by to handle the case of a single variable
Fixed plot.irt.poly so that it communicates correctly with irt2latex for multiple factor case
Fixed irt2latex so that it handles results from plotting irt as well as directly from irt.fa
Fixed polychoric so that it returns meaningful tau values in the presence of missing categories (reported by Gabriele Cantaluppi). polychoric was returning incorrect values of correlations and tau for the case of item responses differing in the number of alternatives, or with different minima. Particularly a problem with very small sample sizes. Also generally improved polychoric.
To make the distinction between Principal Components and Rotated Components and Transformed Components, I have labeled them as PCi, RCi and TCi. I said I did that back in 2013 but I did not seem to have done it!
Modified principal so that it will now handle the covar option correctly when finding principal scores.
Modified bi.bars so that both the left and right panel are positive numbers (requested by Lorien Elleman). Also added the ability to label the columns/rows.
Modified omega.diagram so that labels can be applied to the factors.
Added the use option to splitHalf.
Minor tweak to output of fa.extend for the case of extending to one variable.
Modified fa.congruence to consider the case of missing loadings (requested by Emil Ole William Kirkegaard)
Modified documentation to cortest.bartlett to reflect test residual matrices from fa.
Modified alpha to all the specification of n.obs if using correlation matrix. This allows for finding the s.e. of alpha even from a correlation matrix.
Modified cor.wt to handle the problem of missing data
Modified pairs.panels to allow plotting the raw data points but the weighted correlations
Add the ability to do two level hierarchical factor analysis (i.e, taking out n at level 1, m at level 2. Currently we do this with m=1 for omega)
Added a warnings option to the alpha function as requested (repeatedly) by Markus Graf.
Added two experimental measures of unidimensionality/goodness of fit to alpha.
Modified mediate so that it handles moderation as well
Added the diag and upper option to corPlot and cor.plot to allow for blanking out the diagonal or upper off diagonal (Suggested by David Condon).
Added a number of options to plot.residuals to give more helpful displays
Added (improved) error.bars for taking in data from an external source.
Completely rewrote the mediate function to be more general and to have better graphics. Added the ability to have multiple y variables.
Modified principal so that it will now handle the covar option correctly when finding principal scores.
Corrected a bug in alpha (reported by Tamaki Hattori) to correctly find the s.e. of alpha. I was slightly overestimating the correct values.
Fixed a bug in setCor that was given incorrect SE for regressions in the case of std=TRUE (the other stats were correct, but the t and p values of the betas were wrong)
Fixed a bug in error.bars that led to not drawing catseyes when importing using the stats option. Reported by Niall Bolger
Fixed read.clipboard.upper so that it works properly with names=TRUE
Fixed a bug in fa.parallel that would not report just fa=fa in the case of correlation matrices (reported by Aaron Weidman)
Added rot.mat as an output to the fa and principal objects. Cleaned up the fa and principal rotations so they all work.
Added amplitude, mean, and intercept to cosinor
alpha: fixed the warning about reverse keying when keys have been specified (reported by Marc Heerdink)
Fixed the LaTeX bug in the vignettes (I was trying to float htdp instead of htpb) (as requested by CRAN)
Fixed the LaTex bug in df2latex (same problem)
statsBy: Added the ability to find within group and between group covariances instead of just correlations. This also allows the type of correlation to be specified.
mediate: Can now do more general mediation of x -> M -> Y for multiple X variables and multiple M variables.
mediate.diagram and moderate.diagram are now more functional
fa: Added the option to reject weird solutions when doing iterations of factor solutions (suggested by Vencislav Popov)
fa, principal: Improved documentation of principal and fa to explain the com variable.
alpha: Changed the default for check.keys to be FALSE. This will result in a warning saying that you probably want to reverse keys, and tells you how to do so. In response to comments by Oliver John saying that we were making it too easy to find alpha.
dia.arrow: Added the both option to allow curved arrows to have heads at both ends (useful for setCor and mediate diagrams)
plot.irt and plot.poly: Added the option to plot different colors for each item. (requested by Yanna Weisberg).
Explicitly declared all functions input from the stats, graphics, etc. cor R packages.
pairs.panels can now draw multiple plot characters (see the second iris example).
faBy added a free or fixed option to force loadings to the pooled solution (if possible)
Fixed fa so that the bifactor, TargetT, equamax and varimin rotations return correctly (they were returning the unrotated solution). (Reported by Matthias Bellmann.)
describe: changed describe so that it will give (with *) stats for factors (replacing a functionality that disappeared sometime ago) The resolves the problem reported by Alan Kelly.
Fixed a bug in fa2latex so it will boldface "big values" and cut small values instead of one or the other. (Reported by Simon Kiss)
df2latex and fa2latex: corrected the location of font.size to be inside the table to work correctly (Reported by Simon Kiss).
pairs.panels the cor option now works correctly (as specified in the help file) when showing the regression line (it had been reversed)
principal: Fixed principal to be compatible with changes in fa wrt rotations
Fixed schmid and omega to compatible with all rotations
Modified setCor and mediate so that they just read in the relevant variables when finding the covariance matrix.
Added a simulate data option to mediate to handle mediation models based upon covariance matrices.
Added a density option to error.bars.by so that transparency is easier.
Cleaned up faBy to handle missing data
Modified setCor so that variables can be specified by name or number.
Modified scoreOverlap to handle the case of NAs in the correlation matrix.
Added the ability to find and print item by scale correlations (corrected for item overlap and scale reliability) in scoreOverlap.
Added the ID number used for grouping to be the rowname for the output in cosinor (requested by Ashley Kendall)
Added circadian.stats and circadian.F to do basic circadian stats and group comparisons.
Added circadian.sd to work with circadian.mean to allow t-tests.
Added circadian.reliability to find the split half reliability of phases found by cosinor or circadian.phase.
Renamed best.items and best.scales to bestItems and bestScale. Change the call to be more consistent.
Added two new features to pairs.panels. One allows it to handle character variables (which will be translated to numeric levels) as requested by Richard Cotton. Added the ability to specify the number of breaks in the histogram as requested by Philipp Thomas. Also added the ability to change plot character size independently of correlation print size (cex.cor).
Did a drastic rewrite of the pairs.panels function so that is much cleaner (partly in response to a bug reported by Valdar Tammik and a request by Richard Cotten).
treat character as factor in pairs.panels. Requested by Richard Cotten.
Modified dia.arrow and het.diagram to allow control over the gap and label sizes. (Requested by David Condon).
Added a MAR parameter to cor.plot so long labels will fit
Changed the output of the principal function to not give MLE chi square based fit statistics. (Which really did not make sense.)
Added the sim.correlation function to simulate sample correlation matrices from a population correlation matrix.
Added invisible return from fa2latex and df2latex to allow for capture as a file to put into a latex document. (Requested by David Condon.)
Did a major rewrite of fa.parallel to make the structure more clear and to allow for choosing mean or quantile values of the eigen value comparisons.
Modified fa so that it allows (without warning) for smcs in case nf > n.var/2. This case will still report a df problem in terms of fit statistics. It will still throw a warning if using an oblique transformation.
Modified splitHalf so the splits reported are keyed appropriately if some items need to be reversed.
Added the cta.15 function to lock down cta as published by Revelle and Condon. cta is now the development version, cta.15 is the version reported in R and C.
The smc function has been modified for the case of correlation matrices with some missing values. SMCs are found based upon the complete correlations, and alternative estimates are found for those with missing values.
Added warning message to fa, principal, and iclust so that they stop if there are NA values in the correlation matrix.
Fixed mat.regress bug by implementing it as part of setCor.
Fixed faBy to handle the case of bad data more elegantly
Fixed a bug in sim (seemingly introduced when I changed the way of finding simulated correlation matrices. Bug also affected the other sim.functions)
In factor.stats, sometimes the RMSEA limits do not include RMSEA. Rather than reporting non-sensical values, just report NAs.
fixed bug in bestScales that would cause an error if items were missing in the dictionary.
Fixed a bug in fa.plot where it would not handle data without a title.
Fixed a long standing problem in biplot to allow multiple graphs in the same window, and then, discovered and fixed a problem with the margins getting progressively smaller.
Fixed a bug (reported by Isabelle Rivals) in cohen.kappa (just for weighted kappa) when working with raw data with more than 9 categories.
splitHalf failed for n > 32. Fixed. (Thanks to Tom Booth for having the data set that produced this error.)
test for weird correlation matrices failed in the presence of NA fields in scoreOverlap (reported by Lorien Elleman). Fixed.
Added the cor option to fa so that it can find tetrachoric, polychoric, or mixed correlations. Cleaned up the fa code a bit by using switch instead of convoluted if statements.
changed cor.smooth to detect any eigenvalue < 10^-12 This had been 10^-15
Added the Bonett generalization to the Yule coefficient and added confidence intervals as well. (Suggested by Barry Dwight).
Changed tetrachoric and polychoric to use sadmvn from the mnormt package instead of mvtnorm. This seems to result in a speed up of a factor of 2! For complete data, the results match that of lavCor (from lavaan) exactly (or at least to the 3 decimals that lavCor reports) but seem to be four times faster for dichotomous data and 20% faster for polychorics than lavCor.
Added a message to fa if a Heywood case is detected. Suggested by Sagnik Chakravarty.
Added a covariance option to omega and to schmid. Requested by Qingping He.
Modified describe to add a fast option for larger data sets. Added a discussion about how to speed up descriptions for very large data sets.
Added mediate function to estimate mediation and moderation models. (requested by Dan Molden).
Added mediate.diagram to show the results of mediation models.
Added regression.diagram to show regression and set.cor models.
Modified phi.demo to compare phis with tetrachorics and Yules.
Modified fa.parallel to allow it to use tetrachoric, polychoric, or Yule coefficents (similar to the change in fa).
Added a number of if(requireNamespaceGPArotation) GPArotation:: blah blah to make it more compatible with the new namespace checking and error complaints
Modified scoreOverlap to allow for finding correlations from any non-square data matrix (had been limited to case of non-square where the data were outside a 0 1 range).
Added documentation to alpha and scoreOverlap justifying the r.cor statistic.
Corrected cor.ci to fix a capitalized rho (which led to a failure if not using keys. And, then more importantly, fixed it again so that it will work for overlapping scales as advertised.
Corrected r.test to use df = n-3 instead of n-2. Thanks to Bruno Ernande for reporting this.
Corrected the documentation to fa wrt oblique.scores to correctly represent what it does (I had had it backwards) Thanks to Mark Seeto for reporting this.
Added a message to principal if rotation is not correctly specified and defaulting to "none". Suggested by Sagnik Chakravarty.
Fixed describeData so that it doesn't choke on string variables.
Modified the print.psych function so that irt information is printed for every unit rather than every .1 unit. The data are still there in the plot.info object.
Fixed cor.plot to properly add labels to the plot when plotting factor loadings.
Fixed phi.demo to compare phis with tetrachorics and Yules.
Fixed a print problem for fa.parallel
Fixed a rare case in fa.stats for the case of singular matrices
Fixed the way missing is handled in fa so that it actually works! (problem found by Elizabeth Barrett-Cheetham)
Modified error circles to pass just the x and y values to statsBy.
Added the ability to find polychoric correlations within groups in statsBy (requested by Aidan Wright).
Modified cor.ci to allow for cluster overlap corrections.
Added het.diagram to allow for diagrams of heterarchical structures.
Modified dia.arrow to allow for adjusting where the label appears. Applied this fix to fa.diagram.
Modified fa.sort so that it will sort the results of a principal components analysis (requested by Popov Oleg).
Added the outlier function to find Mahalanobis distance measures for a data matrix and flag those cases that are unusual.
Added draw.cor to show how bivariate densities estimates are used in finding tetrachoric correlations.
Added faBy to allow for factor analysis by groups.
Removed inactive calls to the polycor package
Fixed alpha so that it properly returns the average correlation if given a correlation matrix (reported by Franziska Zuber).
A stylistic change in tetrachoric so that it is easier to read.
Fixed two examples in tetrachoric that were impossible
Changed a test condition that was not allowing vector input to be processed in tetrachoric.
Fixed a bug that would lead polychoric to break if there were no cell entries (bad data condition). Missing cells are now given NA. (Reported by Elina Zaonegina.)
Added read.https function
Added the ability to specify group by name or variable location to describeBy, error.bars.by, violinBy, mssd and rmssd.
Added the equamax rotation option in fa. (requested by Sagnik Chakravarty, with solution by Gunter Nickel)
Added the ability to do factor.extensions for omega analyses.
Added some error checking in polychoric for case of bad data (requested by Simon Kiss)
Added the ability to specify the base pch in error.bars.by (requested by Tham Tran)
Modified tetrachoric and multi.cor to use correct = .5 (that is, to make the correction for continuity a variable rather than a logical to do or not do corrections.) In addition, modified multi.cor to do polytomous by dichotomous variables using a generalization of polychoric (polydi).
Modified plot.irt and plot.poly to work together better and to allow specifications of the xlim and ylim across plots.
Added histBy to draw multiple histograms on a single plot
Added scoreOverlap to adjust correlations between overlapping sets of items.
Modified error.bars.by to draw semi-transarent catseyes and to allow for settings of pch, lty, and color for each variable.
Modified fa2latex to boldface abs(loadings) > cut.
Added the varimin rotation from Suitbert Ertl to principal and fa.
fixed print.psych.vss to properly print the minimum eBIC value. (reported by Lorien Elleman)
made x a matrix in mssd to fix a problem of finding mssd of a single variable (reported by Aidan Wright).
fixed con2cat so that it will handle missing data (correctly)
Modified error.bars.by so that it will not break if there is only one case in a group.
Added con2cat to allow making discrete categories from continuous data.
Modified the cats eyes in error.bars so that colors can be specified (requested by Lauren Tindal).
Modified set.cor so that calls by name or location work equally well.
Added the option to change the size of the correlation numbers in cor.plot to reflect the probability values.
Added a default so that cor.ci automatically draws the scaled cor.plot (requested by David Condon).
Changed the structure of the fa return when we have confidence intervals. It has been forming another fa object, but this leads to confusion when sorting.
Modified print.psych.fa.ci so that it handles the revised fa output.
Modified vss so that it doesn't find factor scores on each iteration.
Added the plot.cor.upperLowerCi function to plot the confidence boundaries for correlations (requested by David Condon)
Modified fa.sort so that it will handle confidence intervals for loadings (requested by Jason French).
Modified fa2latex to boldface significant loadings based upon the confidence intervals (requested by Jason French).
Modified score.irt so that it will change the sign to match the direction of the keys (if supplied)
modified error.bars to that x axis labels can be rotated to be vertical (requested by Greg Miller).
Fixed cor.ci so that it handles case of missing correlations (reported by David Condon)
Modified plot.irt so that labels are applied for multiple factor solutions (reported by Jason French).
Modified fa2latex so that it will not blank out h2 and u2 values for cut >0 (suggested by Daniel Zingaro).
A few tweaks to corr.p to be compatible with corr.test
Improved statsBy so that it can properly handle two grouping variables. statsBy also labels the columns in the within object.
Added pairwise names for factor intercorrelations in print.psych.fa.ci.
Two bugs were introduced into corr.test in version 1.4.3. Both fixed. (reported by louis-charles vannier and Clemens Fell)
Fixed print.psych.fa so that it properly sorts the complexities (bug reported by Kai Horstman)
Fixed fa so that if a non-available rotation is requested, a warning is issued.
Fixed alpha for case of 2 variables.
Fixed the naming of correlations in cor.ci
Fixed the empirical confidence intervals for alpha in alpha.
Added best.scale to empirically combine the best items for a scale to predict particular criteria.
Changed the call to best.items to match fa.lookup.
Added a warning for cor.ci if some of the correlations are NA as would be the resulting scales. Will still find the other correlations but be careful.
Added empirical estimates of fit to the fa and fa.stats functions. These are more robust to misfit for matrices that are not positive definite. (Requested by David Condon and Lorien Elleman).
Added these empirical estimates of fit to vss.
Added the nfactors function to do a call to vss and then do nice graphics
Added fa.organize to hand sort factor output by factor number and by item order.
Added item.lookup to organize items by factor and then by item mean.
Added densityBy to show violin plots. Will do this by groups.
Added cats eyes to error.bars and error.bars.by.
Added normal theory confidence intervals to corr.test. (Suggested by Alexander Weiss)
Added the ability to just resample and not simulate in fa.parallel. (Requested by Ashley Kendall.)
Fixed fa.parallel and fa.parallel.poly for case of very low frequency data. We now resample until we get a matrix with no NA values. (Problem reported by Eric Green).
Added a modification to error bars to print a dark circle at the mean. ( Suggested by Jaroslaw Arlet.)
Add significance and confidence intervals to partial.r (Suggested by Elizabeth Schubach).
cor.plot was incorrectly plotting the numbers by col instead of row. Just a problem for asymmetric matrices. Reported by David Condon.
corr.p was not printing correctly. Reported by Robin Beaumount.
Fixed print.psych.fa to cover several weird cases of no x$R2 or stats.df that were causing HDMD package to fail.
Added the epi data set and epi.dictionary of items.
Added bfi.dictionary to show the item contents of the bfi example.
Added BIC, SABIC, RMSEA, and complexity for each factor model in VSS (vss) (requested by David Condon)
added the alias of fa.congruence and fa.stats as calls to factor.congruence and factor.stats in a slow move to consistency
Added fa.lookup to allow for sorted factor loadings with item contents if using a dictionary for the items.
Modified print.psych.fa so that warnings about how factor score correlations are improper are suppressed (as is the printing of the improper correlations).
defined global=TRUE as default in fa.parallel.poly (reported by Koji E. Kosugi)
changed if(max(R2 > 1) to if(any(max(R2 > 1)) in print.psych.fa (reported by David Condon)
Fixed the se calculation in describe which had been incorrectly changed in version 1.4.1. Thanks to Terry Jorgensen and others for reporting this.
Fixed fa.sort so the that communalities and complexities are sorted as well
Substantial improvements in speed through the use of the parallel package and some coding improvements.
Added parallel processing (multicores) to polychoric, tetrachoric, and mixed.cor. With 2 cores this saves 50%, but this is not a linear function of the number of cores: with 4 cores the savings seems to be 67% and with 8 cores 75%. Requires the parallel package which seems to be supplied with core R. Apparently this benefit will not help those running PCs which don't implement forking.
Added parallel processing using multicores to fa when calculating confidence intervals. Basically changed the interation loop to an lapply and then made that a mclapply.
Added parallel processing using multicores to fa.parallel, cor.ci and alpha (for the empirical bootstrap).
Yet another speed improvement by introducing tableF: a cut down/speeded up version of table. This works only for tabulating two integer vectors, but by reducing error checking, is at least twice as fast. This is used in particular in polychoric and tetrachoric, and therefore in mixed.cor as well.
Changed fa so that confidence intervals are raw rather than fisherz transformed before averaging.
Modified ICLUST to increase speed by about 50% by doing the smcs once instead of every iteration. This does not use multicores so the savings will be for PCs as well as OS X and Unix. 50% of time is now spent doing matrix multiplication – can we speed this up by not rescoring all clusters each time, but rather doing incremental changes?
Minor tweaks to increase speed when creating lists and matrices.
Added omega2latex and modified fa2latex to allow not printing small values.
Added ICC2latex.
Changed score.items to scoreItems in a continuing switch to camelCase. score.items will eventually be deprecated unless the entire score. functions are integrated somehow.
Added option to sim.congeneric (and others) to do multiple simulations with same theta values. (Actually, it was there all along. Just set.seed(some value) before each run).
A minor tweak to r.test to better report the call in the case of a correlated correlation.
A minor tweak to corr.test so that if all the sample sizes are the same, just report one number.
Reduced the number of executable examples to speed up compilation. The examples are just commented out. Worth running for understanding.
Added the Gleser data set for an example of generalizability theory.
Added the ability to find scores automatically in omega.
Added the ability data set (just the iqitems data set scored for correct/incorrect). This is more convenient for demonstrations.
Added the check option in describe (turning it off increases speed drastically, but at the risk of not detecting logical or factor data).
Added the describeData helper to do a quick pass to list the data types in a data.frame.
Rewrote the summary function to use switch rather than a series of ifs.
Added a parameter to factor.scores to allow the data and a correlation matrix to both be passed for scoring.
Added Structure as an output for principal (requested by Emeka Egbuna)
Added best.items and lookup as two helper functions.
sim.cor.poly – fixed so that it actually works.
Fixed alpha so that it will properly reverse key if given correlations. (Not sure when this stopped working.)
Corrected the sign of at.rest in the TA scoring key for the msq.Rd file
Corrected reverse.code to not add a constant to positive items (reported by Jian Jin)
Perhaps finally fixed pairs.panels so it will not change options for graphics window.
Fixed polychoric so that it will not blow up if the number of response alternatives are not equal. (Reported by Jeanette Lim )
Fixed fa.poly so that the scores are calculated based upon the polychoric correlations rather than the Pearson correlations (based partly on trying to solve another problem reported by Jeanette Lim)
Fixed scoreItems for the case of an unnamed keys vector (reported by Sara Weston)
Added a brief wrapper function (corFiml) to call FIML procedures taken from lavaan. Inspired by some code from Joshua Wiley and with the help of Ashley Brown. The lavaan functions necessary for this are not exported from lavaan and hence have been partially moved (with adaptations) to psych.
Completely modified guttman to use the splitHalf function. This provides much more accurate estimates of the greatest lower bound.
Added the splitHalf function to find all possible split halves of scales of up to 16 items, and to sample repeatedly random splits for more than 16 items.
Modified plot.parallel.poly and fa.parallel.poly to allow choice in plotting between both, fa, and pc. (Matches the fa.parallel function).
Added the weight option to fa.poly, polychoric, tetrachoric, and mixed cor. This allows for cases to be given unequal weights. Requested by Fabricio Flalho.
Enhanced the documentation for irt.fa and plot.irt to explain when to use plot.irt vs. plot.poly (Requested by David Condon).
Enhanced cor2latex with the help of Davide Morselli. Now can find correlations directly and can also (if desired) show probability stars.
Added standard errors and 95% confidence boundaries to alpha and score.items (suggested by Doug Lawson)
Minor tweaks to irt2latex and fa2latex.
Corrected the documentation to fa to specify that it allows targetT and targetQ rotation.
Corrected kaiser to return as class "psych" and "fa" so that solutions can be used in fa.diagram. (Reported by Gouri Shankar Mishra)
Corrected alpha so that it reports r.drop in the case of working from a correlation matrix. (Reported by Nicolas Hubner)
A serious speedup to tetrachoric and polychoric was initiated with the help of Jason French. The increase in speed is roughly 1- (nc-1)^2/nc^2 where nc is the number of categories. Thus, for tetrachorics where nc=2, this is a 75% reduction, whereas for polychorics of 6 item responses this is just a 30% reduction.
Added the ability to rotate the numbers in cor.plot by specifying srt. (Thanks to a suggestion on Stack OverFlow by shujaa).
Added the use parameter to the fa function to allow for other than pairwise correlations.
Added an invisible return from the print.psych.fa function to include the variances accounted for and the cumulative variance accounted for. (Requested by Eric Green).
Fixed a bug in print.psych.fa so that it did not always work with the HDMD package.
Modified fa.parallel.poly so that the progress bar is not called during the tetrachorics or polychoric simulations.
print.psych.fa did not work for covariance input when adding the complexity option.
Added Yule2phi and Yule2tetra to convert Yule Q coefficients to phis or tetrachorics. Completely rewrote Yule.inv and Yule2phi.
Added sim.poly.mat to generate polytomous or dichotomous items given a particular item structure.
Modified describeBy to allow specification of digits in the matrix output.
Modified multi.hist to allow one more control over graphic output. (Requested by Lars Carlsen).
added iclust.sort to documentation of iclust.
Added label option to fa2latex (requested by Robert Carlisie)
Added sanitize.latex function to all the latex functions so that they properly escape underscore and ampersand.
Added method option to scatter.hist to allow for other correlation methods
Modified tetrachoric and polychoric so that (0,1) and (1,2) data can be mixed together. tetrachoric and polychoric agree for dichotomous data (as they should)
Added an option to the mixed.cor function to allow for spearman or kendall correlations.
Added Hoffman's complexity index for the factor analyses output
Modified make.keys to allow addressing items by name rather than location.
Modified alpha to allow for specify items to reverse by name or location.
Modified the msq data set Rd file to address by name rather than location
Minor tweaks to fa.plot and cluster.plot to allow more graphic control
modify structure diagram so that the error arrows are to the left for lr printing
Modified fa.plot and cluster.plot to allow positioning and sizing of item labels for two dimensional plots.
Made sure all usage and example lines are less than 90 characters.
Added the ability to calculate factors scores in fa.poly.
Significant improvement in biplot.psych to allow much greater control over plotting.
Added cor.ci to find bootstrapped confidence intervals for raw and synthetic correlations.
Added group reliability calculation for omega.
return (invisibly) the eigen values of the factors and components from scree.
Added unit weighted multiple correlation and unit weighted set correlation to set.cor.
Added test.all to allow for testing whether other packages work (or at least their examples) when psych is loaded. (Can be used for any set of packages.)
Added the ability to simulate as well as resample dichotomous data in fa.parallel.poly (requested by Balal Izanloo).
Drop (with a warning) items that have no variance in alpha (requested by Eric Green). This feature has also been added to score.items.
Modified upperLower so that the rownames come from the lower diagonal matrix, and the column names from the upper diagonal matrix.
fixed statsBy so that it correctly reports the pooled values (reported by John Rauthman)
corrected a problem with fa that was leading to errors when handling very non-positive matrices
fixed a bug in Yule.inv that was leading to impossible values being reported
fixed the way that biplot.psych returns the op$mfrow parameter
applied the same fix to pairs.panels to perhaps fix the recurring problem
Fixed fa.parallel to not break if just principal components output is requested
Identified and fixed bug in mixed.cor for case of improper data.
Identified and fixed? bug in mixed.cor so it will treat randomly mixed data
Corrected corr.test and cor.p to properly pass the adjust option (reported by David Weisman and Russell Pierce)
Corrected error.bars, error.bars.by and errorCircles to use n-1 for df in the call to qt. Modified the documentation to point out that we are using the t-distribution. (Thanks to Trevor Dodds)
Fixed fa.parallel so that it correctly prints out the legend for the case of fa="pc" or fa="fa" (reported by Andrew Hsiao and others)
Corrected the way confidence intervals were found in fa.poly
Correctly name the scored variables in case of impute = 'none' (reported by David Condon
Complains, rather than blows up when doing irt.fa and plot.poly for Heywood cases.
Added column names to factor scores.
Corrected the way that data are randomized in the fa.parallel.poly function so that each variable retains (roughly) the original difficulty. Also did this for fa.parallel. Reported by Balal Izanloo.
Fixed (?) progressBar so that it does not freeze the R.Gui for Mac.
Cleaned up help files so that lines don't over run the page in the pdf version. (Suggested by Jennie Miller). Still seems to be a problem.
Fixed factor.stats so that it does not complain when finding scores for principal
Fixed fa2latex so that the apa option correctly puts in the last cr
Added ECV to omega to give an estimate of unidimensionality
Add a parameter to set.cor and mat.regress to allow for square data matrices (suggested by Thomas Richardson).
Changed the default in principal to find scores (scores=TRUE). This makes it compatible with fa.
Fixed principal so that it labels the components as rotated or transformed (as documented and previously implemented. Somehow this feature went away.
Fixed fa so that specifying SMC=FALSE will replace the diagonals with 1s instead of 0s. (Seems to have been an error introduced when the SMC= a vector option was introduced)
Minor bug fix to sim.omega so that it will not choke on badly structured sem commands.
Added fa2irt to convert factor analysis output to Item Response Theory type output. (Basically doing what irt.fa already does, but from a separate factor analysis.
Added fa.extend to make factor extension easier to do.
Added an overall goodness of fit tests for fa.extension applied to the extension variables.
Cleaned up factor.stats to make it do fm="minchi" more readily.
Added two trivial helper functions r2d and d2r to convert from correlations to effect sizes and back.
added the von Neumann test of successive differences (mssd and rmssd). This is useful for studying within subject changes in mood.
Changed the help files for test.psych, irt.fa, iqitems to dontrun the examples (cutting down time to compile and test the functions)
fix the output in the example for fa. Fixed by noticing that:
in fa, fm="pa" did not return labels for the factors.
in statsBy, fixed a problem if the by returned NULL values for some group values
Finally fixed bug in describeBy for the case of NULL categories for the by variable. (Reported by Nipa Phojanamongkolkij).
fixed bug in smc for case of missing values of diag(smc)
added an error message to cor.smooth for the case of NA eigen values
Changed the random seed in overview.rnw for the example of omegaSem to 17 (from 42). The previous seed was generating a bad omega solution which in turn caused sem to fail.
fixed the call function in ICLUST to properly report the parameters being called.
Added the rangeCorrection function to correct for range restriction.
Added a new factor method (minchi) to the fa function. This weights residual correlations by the pairwise sample size. The minimum weighted residual is then found. This will give OLS (minres) solutions for the case of equal sample sizes, but slightly better (weighted) fits for unequal sample sizes. Most appropriate for the Massively Missing Completely at Random structure of SAPA data.
Added a measure of the Kaiser-Meyer-Olkin index of factoring adequacy
None yet
Updated the overview vignette
Added the ability to do target rotations (TargetQ) in omega.
Improved documentation of fa to explain how oblique SS are found.
A better set of items is used for the iq example (iqitems).
add CVE estimate to fa (see Reise 2010) (had already done this for omega).
Changed ICLUST.sort so that it can sort the loadings of principal (requested by Gudmundur Arnkelsson and then print them with the number of digits requested in the print command.
Modified score.irt so that subjects who miss all items or pass all items are given an estimate based upon the (product) of the difficulty of the items they miss (pass) and then adjusted based upon half the quantile difference from 0 (if they miss all items) and 100 (if they pass all items).
Modified sim.omega to allow for specifying a general factor. This allows for tests of not just the bias in the case of no general factor, but also the ability to detect a general factor. Also modified it to include calls to omegaSem.
modified iclust.diagram so that cluster names can be specified rather than all start with C1 ... Cn (requested by Michael Kubovy)
Slightly improved the documentation for r.test so it is clear which correlation is which.
Modified iclust so that the fit statistic is based upon the off-diagonal elements unless otherwise specified. Use diagonal =TRUE to get fits matching previous analyses.
Modified the print.psych.iclust function to print out the Root Mean Square Residual correlation. (It was previously reported in iclust output, but not printed.)
fixed a bug in score.items, such that if the number of subjects is equal to the number of items, no scores were returned. Reported by Jeromy Anglim.
Modified factor.stats so that rare condition of an exact fit in omega does not lead to an error.
Fixed plot.irt so that item labels are correct
Modified omegaSem to pass n.obs to sem
Fixed df2latex so that it will on text variables as well as numeric variables
Improved statsBy to find within and between group correlations.
Added a data set, withinBetween, to graphically show the issue of multilevel correlations.
Added a simulation function to generate multilevel data (sim.multilevel)
Added a function (irt.responses) to plot responses as a function of the latent score for multiple choice alternatives.
Modified progressBar so that it only shows dots when using the console, but not when using a text file for output (i.e., with Sweave).
fixed plot.poly and plot.irt so that multiple scales for one factor can be plotted.
Slight modification to cor.plot to allow for better control of multiple plots on the same page.
Added the ability to put values into the cor.plot output (numbers=TRUE)
Added a function (lowerUpper) to combine two symmetric matrices and output one as the above diagonal entries, and the other as the below diagonal entries of a square matrix.
Added a function (topBottom) to act like headTail but to do it without ellipses.
In a gradual switch to camelCase, changed the name of headtail to headTail, and describe.by to describeBy. Warning messages are issued about the change.
Added a function, statsBy, to find summary statistics (means, sample sizes, standard deviations) by a grouping variable. Similar to describeBy but somewhat easier to use. Also will report (as an option) the pooled within group correlations.
Added a function to find sample size weighted correlations, cor.wt. Useful for SAPA types of analysis.
Modified superMatrix to take list input so that more than two matrices can be combined at once (suggested by David Condon).
Added errorCircles, a greatly improved version of error.crosses.
Added the option to specify starting communalities in the fa function (requested by David Gosar).
cor.plot was actually switching row and colnames (but since it is mainly used on symmetric matrices, this was not a problem). Fixed.
principal failed when handling missing data in data.frame. Fixed. (reported by Neil Stewart)
(Note that 1.2.3 was not released but was replaced with 1.2.4)
fixed ICLUST so that it converts covar matrices to correlations.
Modified pairs.panels so that the method of correlation can be specified (suggested by Carsten Dormann)
added error values to the structure.diagram/lavaan diagram as an option (suggested by José Luis Cañadas )
Changed the oblique.scores option in fa to be FALSE (use the structure the matrix) following a request by Niels Waller.
Fixed the size of the boxes in dia.rect so that omega.diagram and fa.diagram provide reasonable size boxes for long names (requested by Erich Studerus)
modified output of corr.test so that it announces whether or not it is adjusting the probability values for multiple tests.
added the ability to export dot.plot commands without requiring Rgraphviz. (requested by Erich Studerus). Although just implemented for fa.graph, omega.diagram functions can be exported using fa.graph applied to the schmid$sl object from omega.
Added 3 options to the skew and kurtosis functions, and thus to the describe and describe.by functions as well. (Suggested by Bruce Dudek).
Added the ability to plot general factors (ala omega) in the fa.diagram function.
Added the function lowerCor to find correlations (with pairwise deletion as a default) and print out the lowerMat of the results.
A documentation error in r.test was fixed (reported by Nicholas S. Holtzman). The example of Steiger case B was incorrectly ordered and thus led to incorrect results. Specifying by name (rather than location) is now recommended.
In fa, the rotation was not returned, fixed. (reported by Niels Waller)
predict.psych would not work for a single case (reported by Jonathan Williams). Fixed. Also made the documentation a bit clearer to explain the way the factor scores are standardized.
By adding a global-local parameter to mixed.cor it now seems to work on those data sets with badly distributed polytomous data. (Reported by Eric Smith and David Condon.)
Modified scrub to be a bit easier to use (suggested by Josh Wilt)
Added a check in mixed.cor so that if the data are incorrect, announce it first, rather than waiting until late in the process.
Added a progress indicator to particularly slow functions (e.g. tetrachoric, polychoric, biserial). This will show up when using these functions in other functions (such as irt.fa or mixed.cor).
Modified cluster.cor so that in the case of correlations that are NA, it will not find smcs of the raw correlation matrix. In addition, it will impute scale level correlations based upon the average between scale correlation (previously it was reporting an NA). Requested by David Condon.
Added the helper function lowerMat to print the lower submatrix. Added this as an option to the print.psych function.
Modified mixed.cor so that it will automatically (if desired) find the Pearson, polychoric, and tetrachoric correlation matrices and then reorganize the matrix back to the original form.
Minor modifications to tetrachoric and polychoric to give slightly more helpful error messages in the case of missing data and corrections for continuity.
Minor modifications to summary to give cleaner summary stats for fa and irt.fa analyses.
Added a residual and plot.residuals function to plot either qq plots of the residuals, or a cor.plot of the residuals.
Fixed a few references that were working but doing partial matching in print.psych, test.psych, pairs.panels (e.g, in print.psych digit=digits was changed to digits=digits, test.psych nf was changed to nfactors, col was changed to col.smooth in pairs.panels). This had been flagged by linux checking as a note about "partial matching" for a long time. Finally figured out what it meant.
corrected biplot.psych to properly take the xlim and ylim parameters (reported by Andreas M. Brandmaier)
Correction to tetrachoric (and related functions) to return NA in the case of no cases for x or y. (Reported by David Condon)
Fixed fa so that the score.oblique option actually works. (Reported by Jessica Jaynes).
Added various rotation options to principal to match those in fa.
Added "components" as an option to factor score to score principal components using just the component loadings. This is most appropriate when not rotating the components.
Added the Harman.5 (socio-demographic) data set to allow comparisons with SAS.
The way component scores were found had changed from "regression" to "tenBerge" without being documented. Changed the documentation and added the method of scoring as an option. Problem reported by Alexander Weiss.
This is basically a rerelease of version 1.1.10 trying to solve a problem with 32 bit machines and one test that fails. Originally call 1.11.01 but that one still failed the tests.
Added ... as an option in multi.hist (following a suggestion by Rui Barradas)
Added the all=FALSE option to test.psych. This turns off one test to avoid a problem in the testing on Solaris 32 bit machines.
factor.scores and factor.stats were failing in tests for Linux operating systems, but not Macs or PCs. Added some protection against complex eigen values.
fixed problem in irt.scores reported by David Condon
fixed irt.fa so it will properly plot for polytomous items
Modified scrub to be more general.
Added factor structure as an object in the fa function and modified the print.psych.fa function to note that it is the pattern matrix being reported.
Modified the ICLUST output so that the eigenvalues reflect the independent contribution of each cluster.
Added a comment and an option to ICLUST.sort about the order in which clusters are sorted (suggested by Gudmundur Arnkelsson).
Added the rug option to pairs.panels. This will, by default draw a rug underneath the histograms.
Added a covar option to principal to find principal components from covariance matrices as well as correlation matrices.
Cleaned up the objects returned from fa.parallel.poly to more closely match those from fa.parallel.
Added the ability to find and report residual correlations in set.cor. Minor formatting corrections to set.cor and mat.regress.
Fixed a bug in fa such that the oblique.scores option did not work.
fa.parallel.poly was ignoring the fm parameter and was not printing correctly.
Started to modify all calls to sd and mean to use apply to respond to their deprecation for data.frames and matrices
Changed the release numbering system to reflect year and month of release.
Added a parcels function to form item parcels of size 2 or size 3.
Added a jitter option to the factor.plot function.
Added 3 more factor scoring options to the factor.scores function and changed the default scoring option in fa to be the tenBerge
Added the kaiser function to do kaiser normalization.
Added the df2latex function to make LaTeX tables.
Added plot.circular to draw radar and spider plots
Added a short function to create dummy codes (dummy.code)
Added a score.irt function to find IRT based scores.
Added a table of information by attribute level to be (silently) produced in the plot(irt.fa) function.
Added bifactor and biquartimin as rotation options to fa (based on the Jennrich and Bentler 2011 papers).
Implemented an improvement to print.psych function to use switch. (Following a suggestion by Joshua Wiley ). This does not change the functionality, but makes the code easier to debug and to change.
Added an option to alpha to automatically flip items if they seem to be negatively correlated with total score. (Suggested by Jeremy Miles).
Serious modification to cor.plot to allow for colorRamp data. Makes much prettier correlation plots. (Suggested by David Condon and Joshua Wilt).
Modified sim and sim.simplex to create State Trait Auto Regressive simplex structures (suggested by Deepika Anand).
Added a correct option to irt.fa (passes correct to tetrachoric).
Added the ability for omega to be rerun on the same correlation matrix found by previous omega runs. This speeds up analyses of large matrices but particularly that of tetrachoric/polychoric matrices.
Modified headtail so that it will not choke on mixed numeric and string data
Added a smoothing option to the tetachoric correlation function to smooth the resulting matrix if it is not positive definite.
Added cor.smooth to do a principal components based smoothing for correlations.
Added global=TRUE option to polychoric (and tetrachoric) correlations. If global=FALSE, each correlation is found using the pairwise taus. This seems to match John Fox's polycor function. If global=TRUE, the taus are the datawise values. This will differ from polycor in the case of a great deal of missing data. Somewhat faster and perhaps more appropriate given that we are trying to find a global solution.
Added a correction for multiple tests to the p values in the corr.test function and added a new function to just report probabilities corr.p
Fixed a very subtle bug in score.items for the case of impute="none" where some subjects have no non-missing data. This affected the estimate of the average number of items (num.ob.item) per scale. Reported by Ben Schalet.
Fixed bug in factor.stats that gave much too narrow confidence intervals for RMSEA (Thanks to Rick Zinbarg and Ashley Kendall for pointing out the problem).
revised factor extension to properly handle oblique solutions
fa now reports the call correctly – since introduction of iterations it had reported the wrong call
Fixed mixed.cor so that it can handle several special cases.
Perhaps finally fixed pairs.panels so it leaves us with the original parameter settings
Fixed a problem in print.psych such that mat.regress output would not print. (Reported by Ben Schalet.)
Fixed plot.poly so that it correctly plots the ICC results, particularly for negatively keyed items.
Fixed fa so that the covar option works. (It had been killed when iterations were introduced.)
Modified fa, fa.irt, factor.stats, factor.scores so that they handle bad data more gracefully.
Improvements to set.cor to report the canonical correlations and their chi squares. Also added a plot function to set.cor so that it will plot the eigen values of the canonical correllations.
Added the cushny data set to show the original drug data from student.
Resolved why the denomiator df in set.cor does not precisely match the Cohen example. This seems to be because Systat is doing an unbiased estimator and is thus multiplying by (u+v)/v.
Added a polychoric and tetrachoric option to fa.parallel
Added a keys option to plot.psych (plot.poly and plot.irt) to allow for plotting information functions for selected items while retaining the factor loadings from the complete set. (Requested by Ben Schalet.)
Added a summary of the graphic output for irt.fa that includes the average information for each item as well as where the item is most informative. (Suggested by David Condon.)
Integrated the interactions between fa, fa.poly, irt.fa, fa.diagram, plot, etc. to make commands more consistent.
added a factor analysis for tetrachoric or polychoric correlations (fa.poly) that will work without doing an irt analysis using irt.fa.
Added bootstrapped confidence intervals for loadings and correlations in factor analysis (fa), for reliability estimates in omega (omega) and fa.poly.(Originally requested by Steve Miller and Renee Engeln-Maddox for polychoric correlations, extended to the omega case.)
Modified fa.diagram to include graphic output for fa.extension.
Added Cohen's set correlation to the mat.regress function. (Requested by Amanda Uliaszek.)
Changed the order of the parameters in the mat.regress function to be more consistent with the y~x notation of most regression functions.
Added an oblique.scores option to principal. This means that if oblique rotations are done (not the default) then the scores from principal will be oblique as well.
None yet
Factor extension (fa.extension) now can extend an omega as well as a normal factor analysis.
Modified fa so that the factor scores are found for the oblique factors rather than the orthogonal factors. This now produces factor scores that have roughly the same correlations as do the factors. Factor score statistics remain the same. This leads to a different set of factor scores than factanal finds, for it reports the factor scores for the orthogonal factors.
Modified response.frequencies to incorporate a very nice suggestion by Joshua Wiley to identify unique item responses that are not necessarily integers.
Modified fa.extension to correctly find extended oblique loadings.
check added to omega for bizarre case of 1 factor extracted through PC. (This should not be done, for it makes no sense, but nonetheless, it was leading to omega_h > omega_t.) A warning for this condition is now issued and omega_h = omega_t. Reported by Alison Lewis.
Fixed News so that it works as a NEWS.rd file
Added an example to msq
Added a function to perform factor extension (fa.extension)
Added the Dwyer and Gorsuch data sets as examples of fa.extension.
Added lavaan.diagram to the set of diagram functions. Modified to work with lavaan 0.4-7
Added scatter.hist to draw a scatter plot and associated histograms and densities.
Modified score.items so that it reports the number of missing responses for each person for each scale. It had previously reported the total number of missing responses but did not break this down by scale. Requested by Ryne Sherman. Changed the documentation to reflect that missing is calculated.
Modified error.bars.by so that it can not draw lines (lines=FALSE) if desired. (Requested by Ryne Sherman.)
Added to error.bars.Rd to describe how to supply means and standard errrors.
Made various data files available as separate rather than collated. (e.g., Thurstone, Bechtoldt,Holzinger) in the bifactor data set (they remain there as well). (Requested by Michael Friendly.)
Changed VSS.scree and scree to provide a few more options.
Cleaned up circ.tests to be more compatible with rest of psych.
Clean up output from schmid to more match other functions.
Used tools::compactPDF on the vignettes to make them smaller. Also switched to using png instead of pdf for some of the graphics in the vignettes. This reduced the size drastically.
fixed error.bars.by so that it draws the alpha level instead of 2 alpha level error bars.
fixed fa for special case of 3 x 3 matrix with correlations of exactly .5 which would lead to a non-conforming array problem
fixed partial.r to not round output. For rounded output, the print.psych function defaults to 2 digits (reported by David Freedman)
Finally fixed the output of print.psych.fa so that it labels sorted factors independent of cut (Reported by Reinhold Hatzinger )
Eliminated references to polychor in Yule2poly,
fixed read.clipboard.upper so that it will read labels
fixed mat.regress so that it can work with a single predictor
Added mixed.cor to do mixtures of Pearson, polychoric and tetrachoric correlations
Added legends to error.bars.by bar graph
Modified score.items so it will work on correlation matrices as well as raw data.
Improved the plot.irt and plot.poly functions to plot irt output. Called by plot.psych or generic plot call.
Added a trivial little function, bi.bars to do paired histograms
Modified irt.fa to allow for multiple factors
Modified scrub to be a general recoding function
Modified error.bars.by to allow for ... parameters to be passed into axis
Modified diagram so it will dispatch a number of diagram functions (fa, prinicpal, omega, iclust, lavaan).
Added a biplot.psych function to do biplots of fa or principal results.
Added an unbiased estimate of kurtosis in the kurtosi function.
Added a predict function to fa and principal (predict.fa and predict.principal, both called from predict.psych)
Added lavaan.diagram to draw the output from a cfa or sem from the lavaan package.
Added a function to calculate Mardia's test for multivariate normality
Added lengends and improved the examples for pairs.panels.by
Removed the "short" option in score.items
Added the ability to score scales with greatly missing data (for SAPA). Modified the impute option to include = "none"
Added the term KR20 and KR21 to documentation of alpha
Modified pairs.panels so that with the cor option it can draw regressions and give correlations.
Modified the documentation for error.crosses to give a better example.
Added the ability to read rownumbers (variable names?) in first column of read.clipboard.lower and read.clipboard.upper
Tweaked the zlim default for cor.plot to be c(-1,1) rather than c(0,1)
Minor adjustments to Yule and phi.demo to call tetrachoric instead of polycor, thus making the package work with fewer dependencies.
Cleaned up the fa summary output
Fixed dia.curved.arrow to draw one headed curved arrows (as documented)
Fixed a problem with fa.print option sort – Although giving the correct communalities and uniquenesses in the unsorted option, when sorted, it was not sorting h2 or u2.
Added omega total from sem to the omegaSem function.
error.bars.by was ignoring the user supplied ylim. (reported by Helena Chui) Fixed.
Was drawing incorrect sl solution in omega.graph (reported by Mark Difford).fixed.
omega.Rd and schmid.Rd incorrectly referred to fm=ml as fm=mle. Fixed.
Added iclust and vss as an alternative names to ICLUST and vss to be consistent with most naming conventions.
Added the simple utility scrub to do data cleaning.
Added biserial and polyserial to do biserial/polyserial correlations (not to be confused with point biserial)
Modified score.multiple.choice so that it handles missing data without using imputation
Changed the bfi data set to include more subjects and to correctly define the Openness scale.
Modified score.multiple.choice and score.items to report response frequencies for each item (frequencies based upon number of non-missing responses).
Added the burt data set of 11 emotional variables (from Burt, 1915)
Added draw.tetra to draw tetrachoric correlations for fixed tau values (teaching demo)
Added plot.irt and plot.poly to the plot.psych function. These have options to draw the item characteristic function, the item information function, and the test information.
Added irt.fa to do factor analysis of dichotomous or discrete items by first forming a tetrachoric or polychoric correlation matrix. Then convert the results to IRT form.
Added tetrachoric and polychoric functions to find tetrachoric/polychoric correlations as well as item discrimination parameters.
Added omegaSem and omegaFromSem to calculate omega from a confirmatory factor model using John Fox's sem package.
Added (with permission) Bond's Logical Operations Test (BLOT) data set.
sim.irt was not properly handling vectors of item discrimination or guessing parameters. Fixed.
Minor problem in fa.stats that would lead to an error message if chisquare was NaN.
Minor bug in reverse.code for case of 1 variable (why one would want to use 1 the function for one variable is unclear.)
tetrachoric would not apply the correction for continuity when given a table input but would when given normal data. Fixed.
Added a few parameters to fa.parallel to a) set the y label and b) to show or not show a legend. Useful for multipanel displays.
Minor improvements to fa.diagram to allow for input of factor loadings and intercorrelations from other functions.
Added sim.omega to simulate multifactorial hierarchical structures as a test (and demonstration) of omega.
Minor changes to the parameter call list for sim.minor to be compatible with sim.omega.
Fixed a problem in ICLUST where it would sometimes not stop at the right number of clusters for n.clus = 1
Fixed problem in omega.diagram where it would sometimes use the incorrect line type for hierarchical (non-Schmid-Leiman) diagrams.
Fixed bug in schmid where it would not rotate exactly orthogonal factors and find exactly omega = 0.
Added a warning message that oblique rotation does not work, and does a cluster based rotation instead in the case of exactly orthogonal factors. This is just a problem for artificial data sets.
History (Both of these are probably due to a problem in GPArotation which has since been fixed.)
Rearranged the order of parameters for fa so that it is easier to call.
Added various sem stats to fa output (e.g. RMSEA, BIC, rms, crms). These are also provided for omega output.
Add fit statistics to omega for g solution only.
Added a measure of general factor adequacy in Omega. This is just the percent of common variance that is general factor variance for each item. High values and low variances suggest a more meaningful estimate than lower average values or higher variances.
Various clean up of output of ICLUST and iclust.diagram. Added color option for positive and negative loadings.
Changed ICC to work on complete data (i.e., using na.omit) as the default (following a question) by Ross Culloch
Changed pairs.panels to allow for color choice of histogram (requested by Elaine Kuo)
Changed describe to handle numeric but annotated vectors (suggested by Soeren Vogel)
Clarified documentation of cohen.kappa and allowed for using explicitly non-numerical categories (as suggested by Peter Ehlers and Scot McNary)
Increased the usefulness of error.bars to allow for input of x values, y values, and confidence interval. Also added the ability to use 1 standard deviation.
Minor change to print.psych.fa to round before using formatting.
Removed the cut <- min(cut,max(abs(load))/2) following a request by Reinhold Hatzinger to be consisent with factanal
Added an option to omega to allow for analysis of data sets from just the factor loadings and correlation matrices.
Add a scree function to replace VSS.scree
Fixed a bug in pairs.panels for the case of not smoothing but with data ellipses (reported by Hirayuki Sato)
Fixed a serious bug in omega for the case of non-positive manifold items (e.g., personality items). The sign of the group factors was not correctly being set which would lead to incorrect estimates of the multiple R2.
Fixed bug in ICLUST.diagram where it would draw on top of other figures in rare cases.
Fix bug in ICLUST for the path coefficients – in the case of negative loadings, some of these were much too small.
Correctly color the paths in omega.diagram.
Fixed print.psych.fa to correctly showing item numbers for sorted output.
Corrected tetrachor so it would not inappropriately claim that some cells had no data (this did not affect the correlations, just gave an inappropriate warning).
Added within option for error.bars and error.bars.by to allow for within subject error estimates.
Reformatted this file (NEWS) so that the command (e.g.,) news(query=Version > "1.0.85",package="psych") provides nice information.
Added more detail to the overview vignette to discuss how to find omega.
Added the block.random function to block randomize for experimental designs.
Corrected schmid (and thus, omega) so that it finds orthogonal factors before rotating them obliquely. This corrected an error introduced in version 1.0.86 when the default rotation for fa was switched to be oblimin.
correct=FALSE option in tetrachor actually works
error.bars (and error.bars.by) would not properly draw < 3 variables
Added logit, logistic and logistic.grm functions to help teach about IRT.
Added the ability to factor analyze covariance matrices (with resulting loadings of covariances rather than correlations) (suggested by Andreas Moeltner)
Added two new functions to estimate the greatest lower bound (glb.algebraic, written by Andreas Moeltner) and glb.fa.
Added read.clipboard.fwf to allow reading of fixed width files from clipboard. (Originally this required adding a patch to the read.fwf function but that was subsequentally fixed.)
Minor adjustment to ICLUST.cluster so that it will produce fewer clusters (if desired) than it thinks are appropriate. That is, if n.clus is less than the appropriate solution, a warning is issued, but n.clus clusters are found.
Changed the default rotation in fa to be oblimin. (Subsequently discovered that this introduced a bug into schmid and therefore omega.)
Added the glb.communality function to calculate the greatest lower bound by using factor communality estimates (based upon correspondece with Andreas Moeltner, although his algorithm (glb.algebraic, which is now included) which uses the Rcsdp provides slightly different estimates).
added names to communalities in principal (requested by Gumundur Arnkelsson).
Minor change to headtail
Added the sim.rasch function to simulate 1 parameter IRT Rasch models.
Added the data sets bock and income. bock is the source of the lsat data,income is US family income from 2008.
Seriously modified the estimation of Cohen's Kappa in the wkappa function and added a new function: cohen.kappa to calculate kappa from raw data or from similarity matrices. Also calculates the variances following Fleiss, Cohen, and Blashfield, and then find the confidence limits.
Modified the sim.minor function to simulate correlation matrices with nfact major factors and nvar/2 minor factors with larger number of variables.
Fixed print.psych.fa to give correct communalities for sorted factors (had been not sorting the communalities, although it had the factor loadings).
Minor change to comorbidity to clean up output (using print.psych)
Fixed guttman to correctly estimate lambda 4 (error pointed out by Andreas Moeltner).
Modified the output from fa.parallel to print summary statistics. Also added an option to use smcs as communality estimates.
Added the sim.minor function to generate major and minor factor structures.
Modified fa to use first derivatives based upon the model rather than just empirically derived. This results in an impressive speed improvement.
Continued to improve the dia.x functions. Labeled the entire set of functions diagram (for easier search).
Improved documentation for the dia functions to show the power of structure diagrams
Improved structure.diagram to allow graphs to go left to right (default) or bottom to top (which means causal arrows go left to right)
Added the affect data set which includes data from two experimental studies of affect.
Added some features from pairs to allow group coloring in pairs.panels
Modified factor.parallel to allow for other factoring models (including minres)
Modified fa to give eigenvalues of the correlation matrix, and of the common factor space.
Modified pairs.panels so it can plot even if the correlations are NA
Added a legend and improved the output of cor.plot.
Having learned how to use strwidth, have cleaned up the dia.rect, fa.diagram, structure.diagram, and iclust.diagram functions.
Changed the way fa.parallel handles correlation matrices if the number of subjects is not specified. Rather than issuing a warning and proceeding as if n.obs = nvars, it now assumes n.obs =100.
Improved documentation for fa to explain minimum correlation of factor score estimates.
Fixed a problem with printing of the principal (components) results
Fixed bug in the print.psych.vss function that returned the wrong value (although identifying the correct number) for the Velicer MAP criterion. (Reported by Maike Luhmann).
Added Suggests Rgraphviz to DESCRIPTION file to get around CRAN test for 2.10 (as I should have done according to documentation for packages). Removed all tests of functions which require RGraphviz for 2.9.2 because Rgraphviz does not work on SnowLeopard.
Improved the output for mat.regress to allow print() and summary().
Corrected problem in describe.by to properly identify the matrix output
Corrected bug to fa for case of oblique and negatively correlated factors (reported by Erich Studerus).
Minor improvements to the dia.x functions
Fixed the formatting of the news file to follow the appropriate format
Removed the old supplementary manual from the docs folder to make for a cleaner installation
Improved the documentation for the dia.X functions and made some of the calls simpler.
Added the ability to draw "diagrams" to replace the functions that required Rgraphviz. These are path diagrams with rectangles for observed variables and ellipses for latent variables. A set of graphic functions for drawing diagrams includes dia.ellipse, dia.rect, dia.arrow, dia.curve and dia.self. This allows the function omega and ICLUST to draw structural diagrams without using Rgraphviz.
The documentation for these new functions is a bit less than desirable and the dia primitives will be modified slightly for the next release.
added the Schmid data set which includes the original Schmid-Leiman correlation matrix as well as data from Chen, West, and Sousa (2006).
modified geometric.mean and harmonic.mean to allow for missing data.
Cleaned up describe.by so that it can handle matrix output of multiple grouping variables and then print correctly.
Modified factor.stats to allow for statistics even if the matrix is very singular (i.e., N var >> n.obs).
Corrected bug to ICLUST introduced in version 74 that can, in some conditions, produce negative betas.
Fixed an error in printing communalities for rotated factor solutions reported by Tamaki Hattori. Also resolved a problem of extra lines appearing in error.bars. Minor revision to polar.
Finally resolved the problem of not passing the CRAN tests for Mac (even though developed on a Mac, the test package assumes no other packages).
Reinstituted reporting communalities and uniquenesses for principal components and for factor analysis. This feature had been dropped sometime in the past and reported missing by Ista Zahn.
Added two new data sets: Harman.Holzinger and Harman.Burt. The former is used by Peter Bentler as an example of alternative methods for estimating reliability, the latter is a nice example of what happens if the correlation matrix is singular. smc estimates are inappropriate in this case and need to be checked.
Added a weighting option to ICLUST. Weighted beta finds the average between cluster correlation and takes the ratio of the n ^2 * average between cluster correlation to the total variance of the pooled clusters as a measure of the general factor. That is, Unweighted beta calculates beta based upon the correlation between two clusters, corrected for test length using the Spearman-Brown prophecy formala, while weighted beta finds the average interitem correlation between the items within two clusters and then finds beta from this. That is, for two clusters A and B of size N and M with between average correlation rb, weighted beta is (N+M)^2 rb/(Va +Vb + 2Cab). Raw (unweighted) beta is 2rab/(1+rab) where rab = Cab/sqrt(VaVb). Weighted beta seems a more appropriate estimate and is now the default. Unweighted beta is still available for consistency with prior versions. (Unfortunately, in doing this, a bug was introduced that could produce negative betas. Fixed in Version 78).
Modified the fa function to include maximimum likelihood factor analysis.
Added gls fit option to fa
Added mat.sort to sort correlation matrices by factor loadings.
Revised the overview vignette to make more readable.
Fixed (I hope) the documentation error that has been preventing being loaded onto CRAN for Mac.
Added factor stats to omega
Minor fix to print.psych.omega to correct printing error introduced in .71
Combined principal axis, unweighted least squares (minres) and weighted least squares into one combined function (fa). Eventually the three other functions (factor.pa,factor.wls, and factor.minres) will be phased out. Added more rotation and transformation options (basically by making calls to the GPArotation package).
factor.stats now reports the factor weights matrix for factor scores.
A new factor analysis function factor.wls will give weighted least squares (wls).
Omega will now give an estimate for 1 factor and a message about how there are alternatives for 2 factors.
Fixed a bug in print.psych.omega so that it will now report the item numbers correctly when sorting.
Fixed call in phi.demo so that it checks if the polycor package is installed. This was causing problems in the Mac version on CRAN.
Added factor.stats function to report a variety of statistics for factors or components. In particular, added measures of factorial adequacy. factor.stats is now called by the factor.pa, factor.minres, and principal, so it is easier to adjust all three of these functions.
Added a new rotation option to target.rot – if keys are not specified, it will rotate to a cluster solution.
minor tweak to sim.hierarchical to make it consistent with the Jensen and Weng paper
added to omega to calculate omege_infinity
corrected bug in principal for case of singular matrices (reported by Gudmundur Arnkelsson)
corrected bug in print.psych so that it now sorts by factor loadings correctly
added cor.plot to graphically display correlation structures
modified cosinor to do regression as well as optimization for fits
minor correction to ICLUST to catch non-invertible matrices
Added method="minres" to do minimal residual (ULS) factor analysis to the factor.pa function.
Added the target.rot function to do targeted rotation.
Added the sim.anova function to simulate 3 way ANOVA or linear models.
Minor cleanups of code in other functions
Have included the vignettes in the package
Serious modifications to mat.regress to allow for raw data and covariances to be used.
Added corrections for item overlap to use estimated communalities of items to alpha, cluster.cor, etc.
Added estimates of G6+ (Modified Guttman 6) reliabililty to alpha, score.items
Introduced two package vignettes (overview and psych_for_sem)
minor modification to ICLUST with respect to labeling
addition to factor.congruence to allow for comparing omega solutions with factor solutions
modified describe to allow for descriptions of categorical variables.
Added the function alpha to find Cronbach's alpha and Guttman's Lambda 6 for total scales as well as with single item deleted.
Minor enhancements to the structure.graph and structure.sem functions. Further refinements to the psych_for_sem.pdf tutorial.
Added structure.graph and structure.sem to create graphical output and pre-sem commands for sem.
Modified omega.graph to return sem commands for running the sem function from John Fox.
Added Promax to report factor intercorrelations for promax rotations.
bug fix in VSS and omega to allow for promax rotations.
correction to winsor so that it does not sort the vectors of the input data.frame.
bug fix in principal (for problem of oblique rotations)
continued improvements in the print.psych and summary.psych functions
cleaned up sim.congeneric to reduce the amount of output (now handled through print.psych)
added super.matrix function (to combine two matrices – useful for constructing keys and for doing examples in structure.graph)
Improvements to structure graph.
added the structure.graph function
added circadian functions
renamed some functions to make the index order easier to follow.
Introduced the guttman function to estimate the 6 Guttman (1945) estimates (as well as a few others).
Added print.psych and summary.psych generic functions to clean up the output of many of the functions.
Introduced several new functions:
headtail (combines head and tail functions to produce the first N and last M rows of a matrix or dataframe)
error.bars and
error.bars.by plot means with error bars by a grouping variable
omega revised to allow for negatively keyed items. Detects items that should be reversed.
A relatively stable release.
Passed the tests for R 2.6.2
score.multiple.choice provides scores and item statistics for multiple choice items.
circ.sim modified to be more general item simulations
circ.simulation
item.sim
polar converts two dimensional factor loadings to polar coordinates.
A relatively stable release.
A relatively stable release.
Another upgrade.
Passed tests for R 2.5
The first release to CRAN. Prior versions had been tested at NU.
Included pairs.panels, describe, read.clipboard, iclust, vss, principal.
Another release
Another release
The first package version on the personality-project.org repository.
included some of the functions previously in the "useful R" source file.
worked with R 2.1.0