DMAR (version 1.0.0)

ss_power_composite_factorial_anova: Sample Size or Composite Power for a Factorial ANOVA

Description

Determine the necessary per-cell sample size to achieve a desired level of composite statistical power in a balanced factorial analysis of variance, or, given a per-cell sample size, return the realized composite power. Composite power is the probability that every effect named in effects is statistically significant in the same study, the quantity a design must be planned against when its conclusion requires more than one result to hold at once.

Usage

ss_power_composite_factorial_anova(
  factor_levels,
  effects,
  means = NULL,
  sigma = NULL,
  desired_power = 0.85,
  alpha_level = 0.05,
  n_per_cell = NULL
)

Value

A data.frame with term and value columns, as

ss_power_composite_factorial_ancova returns but with

covariate_R2 0 and n_covariates 0. It carries the

dmar_ss_power class for tidy /

glance and the

dmar_composite_power_factorial class for plot().

Arguments

factor_levels

Integer vector of the number of levels of each factor, one entry per factor (each at least 2). A c(2, 3, 2) argument is a 2 by 3 by 2 design.

effects

A non-empty list naming the effects in the composite; see ss_power_composite_factorial_ancova for the format. Each element gives the factors an effect spans and its f or partial_eta_squared, unless means is supplied.

means

Optional array of population cell means (dimensions factor_levels), or a numeric vector of length prod(factor_levels) in array order, from which each effect's Cohen's f is read given sigma. When supplied, plot() draws the mean pattern. See ss_power_composite_factorial_ancova.

sigma

The common within-cell standard deviation of the outcome, required with means and used only there.

desired_power

Desired composite statistical power (default 0.85). Used only when n_per_cell is NULL.

alpha_level

Type I error rate for each individual F test (default 0.05), the per-test rate rather than a rate for the composite event.

n_per_cell

Per-cell sample size, assumed balanced; if supplied, the realized composite power is returned rather than a sample size planned.

Author

Ken Kelley [email protected]

Details

This is the no-covariate case of ss_power_composite_factorial_ancova, named directly. It does not take a covariate; when a covariate belongs in the model, use ss_power_composite_factorial_ancova, which raises every effect's power for the variance the covariate explains. With no covariate the composite is exact to quadrature precision, since the balanced factorial F tests are exactly noncentral F and exactly independent given the shared error estimate.

See ss_power_composite_factorial_ancova for the model, the one-dimensional integral that evaluates the composite over the shared error estimate, and the figure the plot() method draws. Naming one effect reproduces ss_power_factorial_anova exactly.

References

Maxwell, S. E. (2004). The persistence of underpowered studies in psychological research: Causes, consequences, and remedies. Psychological Methods, 9, 147--163.

Maxwell, S. E., Delaney, H. D., & Kelley, K. (2027). Designing experiments and analyzing data: A model comparison perspective (4th ed.). Routledge. (See Chapter 7 on factorial designs and Chapter 3 on statistical power.)

See Also

ss_power_composite_factorial_ancova for the version that admits a covariate; ss_power_factorial_anova for a single effect

Other sample size for power: power_fisher_exact(), ss_aipe_mixed_effects(), ss_power_R2(), ss_power_R2_sensitivity(), ss_power_c(), ss_power_c_ancova(), ss_power_composite_ancova(), ss_power_composite_ancova_2group(), ss_power_composite_anova(), ss_power_composite_factorial_ancova(), ss_power_composite_factorial_ancova_het(), ss_power_composite_sem(), ss_power_contrast(), ss_power_equivalence_c(), ss_power_factorial_ancova(), ss_power_factorial_anova(), ss_power_indirect_effect(), ss_power_mixed_effects(), ss_power_one_way_anova(), ss_power_pcm(), ss_power_r(), ss_power_rc(), ss_power_reg_coef(), ss_power_reg_coef_sensitivity(), ss_power_rm_anova(), ss_power_sc(), ss_power_sem(), ss_power_smd(), ss_power_split_plot_anova()

Other composite power: ss_power_composite_ancova(), ss_power_composite_ancova_2group(), ss_power_composite_anova(), ss_power_composite_factorial_ancova(), ss_power_composite_factorial_ancova_het(), ss_power_composite_sem()

Examples

# A 2 by 3 factorial ANOVA whose conclusion needs both main effects. Plan
# against their composite, not against either one alone.
ss_power_composite_factorial_anova(
  factor_levels = c(2, 3),
  effects = list(list(factors = 1, f = 0.25),
                 list(factors = 2, f = 0.20)),
  desired_power = 0.80)

# Realized composite power at 25 per cell for a main effect and the
# three-way interaction of a 2 by 2 by 2 design.
ss_power_composite_factorial_anova(
  factor_levels = c(2, 2, 2),
  effects = list(list(factors = 1,          f = 0.30, label = "A"),
                 list(factors = c(1, 2, 3), f = 0.25, label = "AxBxC")),
  n_per_cell = 25)

# Naming one effect reproduces the single-effect planner exactly.
ss_power_composite_factorial_anova(
  factor_levels = c(2, 3), effects = list(list(factors = 2, f = 0.25)),
  n_per_cell = 20)
ss_power_factorial_anova(factor_levels = c(2, 3), effect_indices = 2,
                         f = 0.25, n_per_cell = 20)

# The figure of the purported population effect sizes.
plot(ss_power_composite_factorial_anova(
  factor_levels = c(2, 3),
  effects = list(list(factors = 1, f = 0.25),
                 list(factors = 2, f = 0.20)),
  n_per_cell = 30))

# Stating the effects as population cell means with a common within-cell SD.
# plot() then draws the mean pattern, with error bars of one SD.
cell_means <- matrix(c(10, 12, 11,
                       13, 12, 16), nrow = 2, byrow = TRUE)
plot(ss_power_composite_factorial_anova(
  factor_levels = c(2, 3), means = cell_means, sigma = 4,
  effects = list(list(factors = 1, label = "A"),
                 list(factors = 2, label = "B")),
  n_per_cell = 30))