See the methods we have added, the tasks we have made easier and the issues we have fixed.
Install through Microsoft Store to receive the public app and its updates. The history labels earlier test releases separately from Microsoft Store releases.
Install QuickPLS Pro for Windows through Microsoft Store. Choose an available access offer and start working with models, set-based methods and saved reports.
New
Install the Microsoft-signed Windows app from the QuickPLS Pro Store listing.
Choose a free Launch Access Pass when available, or a monthly, quarterly or annual subscription through Microsoft Store.
Supported models now open with their results on the diagram. The update also makes HTMT tables easier to scan and shows the QuickPLS Pro name consistently.
Improved
Supported model results now open in Graphical Output. Open Key findings when you want the numerical summary.
HTMT and HTMT+ initially show only the lower half of the matrix for a more compact view. The complete matrix remains available, with explanations when constructs are omitted.
The app window and About screen now show the QuickPLS Pro name. About also displays the version, so you can identify the release you are using.
An extra option when structural CB-SEM fitting stalls
Improved convergence gives basic structural CB-SEM another way to fit continuous linear models at the requested precision. It works with covariances alone or with covariances and means; choose Legacy when reproducing an earlier calculation.
Improved
Choose Improved convergence in the basic structural CB-SEM workflow to use the updated solver while keeping your requested tolerance, the precision required for the calculation.
Choose Legacy when you need to reproduce earlier calculations with their original numerical behaviour.
The first use in an existing project creates a protected copy. The original project remains available for use in earlier releases.
Scope
The update applies to basic point estimates. Bootstrap uncertainty, uncertainty for structural effects, user-defined constraints and parameter formulas continue to use their existing calculations.
If a model still fails to converge, the app reports the failure and explains the diagnostic result.
Move from calibration and truth tables to separate outcomes, specialist set methods and reports linked to saved calculations. This release adds tools for several different research questions.
New
Run separate outcomes with shared conditions, then compare the results. Each outcome keeps its own analysis choices.
Use the separate Coincidence Analysis workflow to examine the causal structures and alternative formulas allowed by its settings.
Add exact-binomial tests to truth-table rows for binary crisp data, with a stated comparison probability and significance level.
Improved
Include calibration choices, necessity results, configurations and robustness results in a report linked to the saved analysis.
Larger fsQCA analyses have capacity options that show an estimate of the work involved and the applicable resource limits.
Scope
The expansion also includes multi-value QCA, temporal transformations and panel diagnostics. Each keeps its own input requirements and interpretation.
Expanded capacity remains subject to method and resource limits. Panel diagnostics describe set fit; they do not estimate causal effects.
Open six teaching projects with invented data and saved calculations. Follow the worked lessons locally, inspect the examples and try the steps in a copy of the project.
New
Practise a guided study and report, outcome comparisons and robustness checks, research choices, multi-value QCA, temporal and panel diagnostics, and workflows with multiple outcomes.
The teaching projects include separate examples for Coincidence Analysis and binomial tests using binary crisp data.
The offline learning pack contains six lessons and nine original practice CSV files.
Improved
Open Tutorials and practice data from either Home or Help Center.
The existing SEM examples remain available with their original data and saved results.
This review update helps you continue an interrupted import and identify the right recovery copy. Existing calculations and saved project formats are preserved.
Fixed
After inspecting an import, you can go back, cancel the choice of a replacement file, and continue with the import you already inspected.
If an import was saved but the display cannot refresh, the message now explains what happened. If the save outcome is uncertain, actions that could repeat the import are disabled.
Long labels and titles now stay within the exported SVG and PNG diagram bounds.
Improved
Recovery choices now show the time and full file location so you can identify each copy before choosing one.
When a dataset or its link to a source prevents an action, the message explains the next steps more clearly.
Clearer truth tables, diagram labels and dataset names
The 1.03.09 and 1.03.10 review releases make everyday tasks easier: inspect a compact truth table, read a long diagram label and rename transformed data without repeating the calculation.
Improved
Begin with the compact fsQCA truth table. Open Full evidence for more detail, or inspect the individual cases that form a configuration.
Use the pointer or keyboard to read the full name when a diagram label is shortened to fit.
Rename a transformed dataset after previewing it without running the calculation again. Changing the transformation or its source still requires a new preview.
Import and CB-SEM failures now have plain-language messages. The original diagnostic remains available in Technical details.
Report prompts now name the condition or outcome that needs your explanation.
Fixed
Opening a project successfully now clears an outdated close notification. Warnings and notification history are preserved.
Explanations about samples and source protection now make clear which copies are editable and which projects contain data only.
Keep datasets separate and read more precise PCA results
This release improves how separate and transformed datasets are saved in a project. It also corrects PCA precision and makes the definitions used in new weighted calculations explicit.
Improved
Save and reopen projects containing only a dataset, or an empty model draft, through the supported workflows.
Add a separate dataset or save transformed data without replacing existing project content. You still choose explicitly when to change the dataset used by a model.
Confidence-interval headings, source labels and missing-data summaries are easier to read.
PDF table exports have improved text wrapping and page breaks.
Fixed
Reopening a saved standalone calculation now restores the exact dataset it used, even if another dataset contains identical values.
PCA numerical precision is corrected. Supported new weighted calculations use an explicit definition of weighted moments, the statistical summaries used in estimation. Historical results keep their recorded definitions.
Fixed the affected steps for showing and recovering copula and post-hoc power results.
Mean replacement fills a missing value with a chosen average. This Windows test release adds it to supported weighted-bootstrap, prediction, interaction-term and conditional-analysis workflows.
New
Use mean replacement in the supported weighted-bootstrap, generated-term and conditional-analysis workflows.
Supported prediction and CVPAT workflows calculate replacement means from the training data only. Prediction is evaluated using observed outcomes.
Improved
Preparation choices, saved settings and report results linked to their source are preserved through reopening and export.
Weighted-bootstrap and prediction results include messages that draw attention to relevant issues and explain when a result is unavailable.
Scope
This unsigned Windows release was distributed through the private repository. It was not a Microsoft Store release.
The planned new pairwise treatment of missing data was deferred and is not included in this release.
QuickPLS