Independent calculations and app workflow

Time-based set calculations and panel summaries checked

These tests check whether a value from an earlier occasion belongs to the right unit, whether event order can be traced to the original rows, and whether panel summaries describe the intended observations. The time-based transformations and the pooled, within-unit and between-cluster summaries are checked separately.

Reviewed 4 November 2026
Specific testsChecks of time-based transformations and panel set fit

What the comparison shows

The temporal tests cover explicit lags, event order, start/end transformations and conservative QCA solutions. The panel tests cover set-fit summaries and an installed-app journey that saves, reopens and exports the report. Optional R comparisons were not run. These results support the listed calculations and workflow, not a general dynamic causal model or unrestricted time-series analysis.

Why must the unit and occasion stay attached to each value?

Panel data contain repeated observations of the same units. A lag is a value taken from an earlier occasion. It must come from the intended unit and time, rather than another unit or a convenient nearby row. An event-order transformation must likewise show which original observations establish the order.

The panel summaries ask different questions. Pooled results use the included observations together, while within-unit and between-cluster summaries use the specified unit and cluster structure. In a country-by-year study, for example, the country can identify the unit and the year the cluster. Those columns affect which observations are compared; they are not just report labels.

Which versions and kinds of evidence are listed?

The two records describe the saved temporal and panel tests, including checks of the calculations and the stated app workflows. The optional comparison with an R reference package was not run for either record.

On a small screen, swipe the table sideways to see every column.

RecordQuickPLS build or result checkedReferenceComparison outcome
Temporal QCA and transformations evidence1.03.01 / Installer D1.03.02See which mathematical and app-workflow checks were performedMathematical reference plus checks of the specified calculations and app workflows
Panel and cluster set-fit evidence1.03.02See which mathematical and app-workflow checks were performedMathematical reference plus checks of the specified calculations and app workflows

Lags, event order and start/end transformations

The 15 calculation tests cover explicit lags, event order and start/end transformations, as well as conservative-only QCA. Conservative solutions use the observed positive configurations without simplifying through unobserved configurations.

The time rules are explicit. An exact-time lag and a chosen number of earlier observations are different operations; a missing exact period is not silently replaced by the previous available observation. Checking the source rows makes that distinction visible when interpreting a derived value.

  • The comparison uses tests of the calculations; no external package was run for this check.
  • It does not establish equality with QCA's special dash/don't-care treatment of temporal configurations or cover every temporal procedure.
What was checkedValue / scope
Calculation-engine tests15

Panel set-fit calculations and the saved report

The 11 calculation tests cover pooled, within-unit and between-cluster summaries. Set fit describes how closely the memberships follow the selected necessity or sufficiency relationship. The installed-app test separately saves, reopens and exports the panel report.

These are complementary checks. The calculation tests examine the summaries, while the report workflow checks whether the selected results can be saved and reopened. One type of check should not be substituted for the other.

  • The optional SetMethods cluster comparison in R was not run.
  • The summaries describe set relationships. They are not estimates of the causal effect of changing a condition.
What was checkedValue / scope
Calculation-engine tests11

How to read the temporal and panel evidence

For a temporal result, inspect the unit, time rule and original rows behind each derived value. A correctly calculated number attached to the wrong occasion would answer the wrong research question. For a panel result, identify which observations contribute to each type of summary before interpreting it.

The saved report checks establish the stated save, reopen and export workflow for the panel example. The calculation counts describe focused tests, not independent longitudinal studies. They do not establish general dynamic causal inference.

  • There is no single decimal-error limit for the page because it combines transformation, calculation and workflow checks. Inspect the expected outcome for the particular operation.

How to inspect the time-based and panel report

The JSON summary identifies the tested versions, types of transformation and panel summary, test counts and recorded workflow results. It also names the source documents.

Repeating a check requires the original observations and the same unit identifiers, time rules, model, settings and software versions. The download reports completed work; it does not include the original datasets or app source code.

Download the comparison results

See the values we checked, the differences we found and the software versions used. The download also identifies the original test files.

About these tests

These are QuickPLS development comparisons. Each record identifies the reference software or mathematical calculation, the tested build and the settings used.

How to understand these comparisons · Find checks for another method