d_effect: General interface for Cohen's d

In MOTE: Effect Size and Confidence Interval Calculator

General interface for Cohen's d

Description

'd_effect()' is a convenience wrapper that will route to the appropriate Cohen's *d* helper function based on the arguments supplied. This allows users to call a single function for different study designs while maintaining backward compatibility with the more specific helpers.

Usage

d_effect(
  m1 = NULL,
  m2 = NULL,
  sd1 = NULL,
  sd2 = NULL,
  u = NULL,
  sig = NULL,
  r = NULL,
  mdiff = NULL,
  sddiff = NULL,
  t_value = NULL,
  z_value = NULL,
  p1 = NULL,
  p2 = NULL,
  n1 = NULL,
  n2 = NULL,
  n = NULL,
  a = 0.05,
  design,
  ...
)

Arguments

m1	Means of the two conditions or measurements.
m2	Means of the two conditions or measurements.
sd1	Standard deviations for the two conditions or measurements.
sd2	Standard deviations for the two conditions or measurements.
u	Population or comparison mean for one‑sample t‑designs, used when 'design = "single_t"'.
sig	Population standard deviation for z-based designs, used when 'design = "z_mean"'.
r	Correlation between the paired measurements (used for repeated-measures designs such as '"dep_t_rm"').
mdiff	Mean difference between paired observations.
sddiff	Standard deviation of the difference scores.
t_value	t statistic value for the test. Used in designs where the effect size is derived directly from a reported t-value (e.g., '"dep_t_diff_t"', '"ind_t_t"', or '"single_t_t"').
z_value	z statistic value for the test. Used in designs where the effect size is derived directly from a reported z-value (e.g., '"z_z"').
p1	Proportion for group one (between 0 and 1), used in the '"prop"' design.
p2	Proportion for group two (between 0 and 1), used in the '"prop"' design.
n1	Sample sizes for the two independent groups (used for independent-groups designs such as '"ind_t"').
n2	Sample sizes for the two independent groups (used for independent-groups designs such as '"ind_t"').
n	Sample size (number of paired observations).
a	Significance level used when computing confidence intervals. Defaults to '0.05'.
design	Character string specifying the study design.
...	Reserved for future arguments and passed on to the underlying helper functions when appropriate.

Details

- '"delta_ind_t"' — independent-groups t-test using the delta effect size, where the SD of group 1 is used as the denominator. Supply 'm1', 'm2', 'sd1', 'sd2', 'n1', and 'n2'. In this case, 'd_effect()' will call [delta.ind.t()] with the same arguments.

- ‘"g_ind_t"' — independent-groups t-test using Hedges’ g, which applies a small-sample correction to the standardized mean difference. Supply 'm1', 'm2', 'sd1', 'sd2', 'n1', and 'n2'. In this case, 'd_effect()' will call [g_ind_t()] with the same arguments.

- '"z_z"' — one-sample z-test effect size where the *z* value is supplied directly along with the sample size 'n'. Supply 'z_value' and 'n'. You may optionally supply 'sig' (population SD) for descriptive reporting. In this case, 'd_effect()' will call [d_z_z()] with the same arguments.

Value

A list with the same structure as returned by the underlying helper function. For the current paired-means case, this is the output of [d_dep_t_avg()], which includes:

• ‘d' – Cohen’s d using the average SD denominator.

• 'dlow', 'dhigh' – lower and upper confidence limits for 'd'.

• Snake_case aliases such as 'd_lower_limit' and 'd_upper_limit'.

• Descriptive statistics (means, SDs, SEs, and their confidence limits) for each group.

Supported designs

- '"dep_t_avg"' — paired/dependent t-test with average SD denominator. Supply 'm1', 'm2', 'sd1', 'sd2', and 'n'. In this case, 'd()' will call [d_dep_t_avg()] with the same arguments.

- '"dep_t_diff"' — paired/dependent t-test using the **SD of the difference scores**. Supply 'mdiff', 'sddiff', and 'n'. In this case, 'd()' will call [d_dep_t_diff()] with the same arguments.

- '"dep_t_diff_t"' — paired/dependent t-test where the *t* value is supplied directly. Supply 't_value' and 'n'. In this case, 'd()' will call [d_dep_t_diff_t()] with the same arguments.

- '"dep_t_rm"' — paired/dependent t-test using the repeated-measures effect size d_{rm}, which adjusts for the correlation between measurements. Supply 'm1', 'm2', 'sd1', 'sd2', 'r', and 'n'. In this case, 'd()' will call [d_dep_t_rm()] with the same arguments.

- '"ind_t"' — independent-groups t-test using the pooled SD (\(d_s\)). Supply 'm1', 'm2', 'sd1', 'sd2', 'n1', and 'n2'. In this case, 'd()' will call [d_ind_t()] with the same arguments.

- '"ind_t_t"' — independent-groups t-test where the *t* value is supplied directly. Supply 't_value', 'n1', and 'n2'. In this case, 'd()' will call [d_ind_t_t()] with the same arguments.

- ‘"g_ind_t"' — independent-groups t-test using Hedges’ g, which applies a small-sample correction to the standardized mean difference. Supply 'm1', 'm2', 'sd1', 'sd2', 'n1', and 'n2'. In this case, 'd_effect()' will call [g_ind_t()] with the same arguments.

- '"single_t"' — one‑sample t‑test effect size using the sample mean, population mean, sample SD, and sample size. Supply 'm1' (sample mean), 'u' (population mean), 'sd1', and 'n'. In this case, 'd()' will call [d_single_t()] with the same arguments.

- '"single_t_t"' — one-sample t-test effect size where the *t* value is supplied directly along with the sample size 'n'. In this case, 'd()' will call [d_single_t_t()] with the same arguments.

- '"prop"' — independent proportions (binary outcome) using a standardized mean difference (SMD) that treats each proportion as the mean of a Bernoulli variable with pooled Bernoulli SD. Supply 'p1', 'p2', 'n1', and 'n2'. In this case, 'd()' will call [d_prop()] with the same arguments.

- ‘"prop_h"' — independent proportions (binary outcome) using Cohen’s \(h\) based on the arcsine-transformed difference between proportions. Supply 'p1', 'p2', 'n1', and 'n2'. In this case, 'd()' will call [h_prop()] with the same arguments.

- '"z_mean"' — one-sample z-test effect size using a known population standard deviation. Supply 'm1' (sample mean), 'u' (population mean), 'sd1' (sample SD, used for descriptive CIs), 'sig' (population SD), and 'n'. In this case, 'd_effect()' will call [d_z_mean()] with the same arguments.

Examples

# Paired/dependent t-test using average SD denominator
# These arguments will route d() to d_dep_t_avg()
d_effect(
  m1 = 5.57, m2 = 4.43,
  sd1 = 1.99, sd2 = 2.88,
  n = 7, a = .05,
  design = "dep_t_avg"
)

# You can also call the helper directly
d_dep_t_avg(
  m1 = 5.57, m2 = 4.43,
  sd1 = 1.99, sd2 = 2.88,
  n = 7, a = .05
)

MOTE documentation built on Dec. 15, 2025, 9:06 a.m.