- Adds additional analyses for reporting to
by_extension_3.R - Adds code for boxplots and violin plots for all cohorts that collected BSID-III data
- Adds
by_extension_3.Randcontourplot_by3.Rscripts to support theby3extension procedure for keygsed2510 - Adds diagnostic info from fitting key
gsed2510 - Reformats to conform to aiR standards
- Major update to create new key
gsed2510. - Updates all scripts to use the
gs1andgl1instrument codes for GSED SF and LF, respectively - Adapts scripts to gsed3 lexicon
- Moves
calculate_DIF_classification()andcalculate_DIF_table()from gsedscripts (0.27.0) to dfine package (0.13.0) - Moves
build_database_fixed.Rfromgsedreadtogsedscriptspackage
- Add some improvements to scripts
- Prepare for reproducibility
-
Adds reporting outline
Rmd/coremodel_aug25.Rmdto document the reporting process for the GSED key -
Adds many scripts to
inst/scripts/phase2directory to help fitting the new phase_1+2 models -
Removes
DUCKDB_LOCALsystem variable from all scripts -
All scripts depends now only on system variables
GSED_PHASE1andGSED_PHASE2 -
Adds
calculate_DIF_classification()to test for DIF by phase and country -
Extend scripts
plot_pass_age.Randplot_pass_dscore.Rwith BSID and apply system variablesGSED_PHASE1andGSED_PHASE2 -
Adds generic script
phase2/fit_core_model.Rwhich replaces293_phase1_long.R,293_phase2_wide.R -
Updates the following phase2 scripts:
- descriptives/plot_pass_age
- models/293_0_ph
-
Adds starting phase2 scripts:
plot_pass_age.R293_phase1_long.R293_phase2_wide.R
-
These scripts replace the private
dmetricpackage by the publicdfinepackage. -
The experimental
lean2class is no longer supported, and removed.
- Adds script
inst/scripts/models-phase1/293_0_test_phase1. This script refits the 293_0 model on phase 1 data using better duplicate removal, duckbd dataset, and removal of inter-rater scores.
- Adds script
inst/scripts/DIF_demo_residuals
- Adds two working scripts
inst/scripts/models/818_6_enhancements.R, a further development of818_6.Rinst/scripts/293_0_BSID_experimental.R, a script to explore a better linking method, using the BSID-III and the 293_0 model. Needs to be checked and - if OK - generalized
- Extends
scripts/DIF_demo.Rwith >2 groups
- Updates
scripts/models/293_0.Rto use DIF tabulation function and explore Yen's Q3 method for exploring dimensionality - Updates
scripts/models/lf_155_0.Rto feature newer functions and to compact code - Adds
scripts/DIF_demo_2groups.Rto shows how to use thedifRpackage in the SMOCC data for testing 2-group DIF with MH and logistic regression methods - Adds
calculate_DIF_table()for testing DIF
-
Substantial update 293_0 script
- better stepwise approach
- use of new functions to reduce copies
- add item deletion of rare categories
- add code to test D-score - logit alignment
- check anchors
- graph item and person fit
- add tests for DIF by cohort
-
Generalizes
make_wide()to provide more records for analysis
- Updates function
update_required_packages()to automate modelling scripts - Adds some dependencies into Suggests
- Adds function
update_required_packages()to automate modelling scripts - Adds function
calculate_administrative()for deriving standard administrative variables - Adds function
make_wide()to combine SF, LF and BSID data from the first observation period onto one record per child - Adds a script
inst/scripts/compare_domain_taus_293.Rto inspect the overlap between three streams of the LF a as way to informally inspect the uni-dimensionality of the LF. Used for the July 2024 version of the phase1 paper. - Adds a scripts
inst/scripts/extract_equate_table.Rto extract the equate table to compare keysgsed1912andgsed2406. Used for the July 2024 version of the phase1 paper.
- Adds
Rmd/20240802_logit.Rmdto demonstrate that logit and D-score scales are consistent for LF and SF, but not for other instruments - Adds new procedure
inst/script/models/by3_extension.Rto link BSID-III under fixing SF/LF tau estimates to those in model 293_0, leading to consistent logit and D-score scale for BSID-III
- Perform updates of model 293_0 using additional LF data for BGD
- Replaces
fit_phase1_healthy_reference.Rbyfit_preliminary_standards.R
- Adds a new fuzzymatching method to
inst/scripts/counts.Randinst/scripts/293_0.Rthat takes only the first LF, SF and BSID before matching. As a result, each row corresponds to a unique subjid (child). This does not use all data, but is much easier to explain and report. The impact of the method on the item difficulty estimates is negligible (r > 0.99).
- Adds script
inst/scripts/fit_phase1_healthy_references.Rto calculate new healthy references for the GSED key - Copies code for creating Dutch references for the GSED key into file
inst/scripts/fit_phase1_dutch_references.R
- Extends script
inst/scripts/fit_models.Ra refit usingdscore 1.8.8 - Adds script
inst/scripts/compare_algorithm.Rto compare models "20221201_remodel/293_0" and "20240601/293_0" usingdscore 1.8.8 - Adds script
inst/scripts/compare_models.Rto compare models "20221201_remodel/293_0" and "20240601/293_0" using thedscore 1.8.8 - Extends script
inst/scripts/fit_phase1_references.Rwith calculation of the healthy subsample references from the GSED key - Extends script
inst/scripts/fit_phase1_references.Rwith calculation of the Dutch references from the GSED key
- Changes path in source scripts for GSED SF and BSID so that it will run from any directory, as long as the R packages dmetric and gsedscript are installed on the user's machine.
- Adds source scripts for document that explain scoring GSED LF and GSED SF
- Make scripts better useable as external call
- Updates modeling scripts
- Updates reports that documents actions taken to solve LF item order problem
- Adds document to log actions taken to solve the LF item order problem
- Updates data reading scripts
assemble_data.Randedit_data.R
- Adds inst/scripts/counts.R to produces basic counts on LF, SF and BSID
- Adds key update and scoring Rmd's for reporting
- Adds scripts/export_gsed_sample.R which draws 10 sample cases
- Moves .txt files to ../export
- Moves .pdf files to ../plots
- Added a
NEWS.mdfile to track changes to the package.