LoginSign Up
← Back to all sparks
S

stdmod

ANALYTICS
Velocity2.5

Standardized Moderation Effect and Its Confidence Interval

github.com/sfcheung/stdmod ↗

A moderation-analysis package now pointing users at its own siblings for the harder work.

moderation-analysisregressionstatisticsr-packagemaintenance
◆Current state
stdmod computes standardized moderation effects in regression, part of a cluster of R packages from the same author covering moderation, mediation and model comparison. Its release feed is a CRAN-announcement format — several entries carry nothing but a version and a link — and its development has slowed markedly, with the substantive feature work sitting back in 2024. The most recent release is documentation rather than code.
◆Where it's heading
The clearest signal is the latest release redirecting users toward betaselectr and manymome for tasks stdmod also covers, on the grounds that those packages handle them more comprehensively. That is a package consciously narrowing its scope within a family rather than competing with its siblings. The earlier feature work — conditional effects at chosen moderator values, R-squared increase reporting, print formatting — reads as a stable core that has since been left alone.
◆Prediction
Expect stdmod to stay in maintenance while the author's newer packages absorb the overlapping functionality, with future releases likely limited to CRAN compliance and documentation.

◆Recent moves

  1. 22d ago

    Docs now steer users to betaselectr and manymome

    Articles and help pages are updated to recommend betaselectr and manymome for several tasks stdmod also covers, on the stated grounds that those packages support them more comprehensively. No code changed, but it is the most informative release in this window: a package explicitly ceding territory to its siblings. It reframes stdmod as the narrower, older tool in a family rather than the entry point.

    View source ↗
  2. 7mo ago

    v0.2.12 at CRAN

    A CRAN release announcement carrying only version links, with the changelog hosted off-feed. No detail about what changed is available here. It ends a two-year gap in releases.

    View source ↗
  3. 2y ago

    Summary printout gains rounding and p-value formatting control

    Improves the printed summary for std_selected() and its bootstrap variant: small numbers round rather than falling into scientific notation, small p-values print in conventional forms, and the number of digits is user-controllable. Presentation work, but for a package whose output goes into write-ups it is the part users interact with most. Bundles a fix to the package-level documentation.

    View source ↗
  4. 2y ago

    R-squared increase reporting and chosen moderator values

    Adds the R-squared increase of the highest order term and its F test to summary output where the model has exactly one such term, and a w_values argument letting users specify the moderator values at which conditional effects are computed. Both address the standard reporting requirements of a moderation analysis. Published seconds before the lower-numbered 0.2.9 entry, so this feed's version order and publication order disagree.

    View source ↗
  5. 3y ago

    v0.2.0.0 at CRAN

    A stable CRAN release announcement with no accompanying notes beyond installation instructions. Nothing about the contents of the release is recorded in the feed.

    View source ↗
  6. 4y ago

    v0.1.7.4 at CRAN

    A stable CRAN release announcement carrying only installation instructions. As with several entries in this feed, the substance of the release is not published here.

    View source ↗