Show distributions of the "evidence" as a function of sample size, evidence scale/power, and true effect size

Source: R/fig_desc_evidence.R

Show distributions of the "evidence" as a function of sample size, evidence scale/power, and true effect size

fig_desc_evidence(
  n = 10,
  q = 7,
  d = NULL,
  letter = "",
  scale_y_on = 1,
  d_scale = 1
)

Arguments

n

sample sizes (single value or vector)
q

evidence scale/power (single value or vector)
d

true effect size (single value or vector)
letter

Label to add to the graph
scale_y_on

How to choose a scale for the y axis; see details
d_scale

Scale multiplier to be applied to effect sizes

Value

Returns the maximum y value (useful for further plotting)

Details

scale_y_on can be one of several values:

  • NULL: scale on the the median maximum density value across distributions

  • 0: scale on the maximum density value across distributions

  • negative real number: the absolute value of this value will be used as the maximum y value

  • positive integer: the index of the distribution whose mode will be used as the maximum

Examples

fig_desc_evidence()