alphaN() gains two methods based on Klauer, Meyer-Grant
& Kellen (2024, Psychonomic Bulletin & Review,
doi:10.3758/s13423-024-02612-2): method = "ES" calibrates
alpha to their effect-size Bayes factor, whose prior centers the
alternative hypothesis on a prespecified effect size, and
method = "moment" calibrates alpha to their moment Bayes
factor, under which effects near zero are a priori implausible. New
arguments de (targeted effect size, default 0.5),
nu, and r control the priors, with defaults
following the paper’s recommendations. Because the moment prior rules
out near-zero effects, the alpha level it implies falls much faster with
n than under JAB.method = "ES", nu = 1, de = 0 with
an explicit r calibrates alpha to the default
(Jeffreys-Zellner-Siow type) Bayes factor of Rouder et al. (2009).alphaN() and JABt() now return correct
results when n is a vector and
method = "robust" or method = "balanced".
Previously, "robust" silently applied the smallest sample
size to every element and "balanced" failed with an
unrelated error.method, a missing
df in JABp(..., z = FALSE), a p
outside (0, 1], a non-positive n or BF, and an
unknown covariate in JAB() (which now lists
the coefficients available in the model).JAB_plot() gained an upper argument,
passed on to the underlying computations for
method = "balanced".alphaN_plot() gained a ylim argument. The
default now covers all four curves; previously the y-axis was fixed to
(0, 0.05), which silently clipped the "balanced" curve for
small Bayes factors.JAB() now determines the sample size via
nobs().?JABp no longer has a placeholder title.JABp()
expects a two-sided p-value.NEWS.md file to track changes to the
package.
Need a high-speed mirror for your open-source project?
Contact our mirror admin team at info@clientvps.com.
This archive is provided as a free public service to the community.
Proudly supported by infrastructure from VPSPulse , RxServers , BuyNumber , UnitVPS , OffshoreName and secure payment technology by ArionPay.