slm_effect_sizes: Compute effect sizes for mass-univariate GLM analysis.

In dfsp-spirit/brainnet: Machine learning on structural neuroimaging data for R

Compute effect sizes for mass-univariate GLM analysis.

Description

Compute effect sizes for mass-univariate GLM analysis.

Usage

slm_effect_sizes(
  X,
  Y,
  predictors,
  output = c("F", "p", "etasq", "partial.etasq", "cohens.f", "rsq", "power")
)

Arguments

X	numerical matrix, the design or model matrix, typically created from the demographics data using model.matrix.
Y	numerical matrix, the target value, typically neuroimaging data
predictors	vector of character strings, the names of the predictors in the model matrix
output	vector of pre-defined character strings, defined what values to return. Leave alone if in doubt.

Value

named list with entries according to the output parameter. By default F= the F value map, p = the uncorrected p value map, etasq = the eta squared value map, , parial.etasq = the partial eta squared value map, rsq = the r squared value map. power = the power of the F test (1 minus Type II error probability) to detect an effect of the computed effect size (see 'cohens.f' entry) given the sample and a significance level of 0.05.

Author(s)

C Ecker, documentation by T Schaefer ″

dfsp-spirit/brainnet documentation built on July 11, 2022, 5:54 a.m.