Compute effect size (Cohen's d / Hedges g)

Calculates the standardized mean difference at each time point and channel, with an exact non-central-t confidence interval (Cumming 2014) and an optional Hedges small-sample bias correction.

Usage

effectSize(
  x,
  condition1 = NULL,
  condition2 = NULL,
  pooled = TRUE,
  correction = c("none", "hedges"),
  conf_level = 0.95
)

Arguments

x

An epoched PhysioExperiment object (4D data).

condition1

Indices for first condition.

condition2

Indices for second condition. If NULL, computes d against zero.

pooled

If TRUE (default for two-sample), uses pooled standard deviation.

correction

"none" for Cohen's d (default), or "hedges" to apply the Hedges bias correction J = 1 - 3/(4*df - 1) to the estimate and interval.

conf_level

Confidence level for the interval (default 0.95).

Value

A list containing:

d

Matrix of standardized mean differences (time x channel)

ci_lower

Lower confidence limit for d (non-central t)

ci_upper

Upper confidence limit for d (non-central t)

correction

The bias correction applied

References

Cumming, G. (2014). The New Statistics. Psychol Sci 25(1):7-29. Hedges, L.V. (1981). J Educ Stat 6(2):107-128.

See also

tTestEpochs() for significance testing, bootstrapCI() for bootstrap confidence intervals, rankBiserial() and cliffsDelta() for rank-based effect sizes.

Examples

# Create example epoched data
set.seed(123)
epochs <- array(rnorm(100 * 4 * 20 * 1), dim = c(100, 4, 20, 1))
pe <- PhysioExperiment(
  assays = list(epoched = epochs),
  samplingRate = 100
)
# Effect size for one-sample
result <- effectSize(pe, condition1 = 1:10)
# Hedges g between conditions
result2 <- effectSize(pe, condition1 = 1:10, condition2 = 11:20,
                      correction = "hedges")