CrucibleIR.Reliability.Stats (CrucibleIR v0.2.1)

View Source

Configuration for statistical testing and analysis.

Controls which statistical tests are run, significance levels, and other analysis parameters.

Fields

  • :tests - List of statistical tests to run (default: [:ttest, :bootstrap])
  • :alpha - Significance level (default: 0.05)
  • :confidence_level - Confidence level for intervals
  • :effect_size_type - Type of effect size to calculate
  • :multiple_testing_correction - Correction method for multiple tests
  • :bootstrap_iterations - Number of bootstrap iterations
  • :options - Additional statistics options

Available Tests

  • :ttest - Student's t-test
  • :bootstrap - Bootstrap resampling
  • :anova - Analysis of variance
  • :mannwhitney - Mann-Whitney U test
  • :wilcoxon - Wilcoxon signed-rank test
  • :kruskal - Kruskal-Wallis test

Effect Size Types

  • :cohens_d - Cohen's d
  • :eta_squared - η² (eta-squared)
  • :omega_squared - ω² (omega-squared)

Summary

Types

correction()
effect_size()
t()
test()

Types

correction()

@type correction() :: :bonferroni | :holm | :fdr | atom()

effect_size()

@type effect_size() :: :cohens_d | :eta_squared | :omega_squared | atom()

t()

@type t() :: %CrucibleIR.Reliability.Stats{
  alpha: float(),
  bootstrap_iterations: pos_integer() | nil,
  confidence_level: float() | nil,
  effect_size_type: effect_size() | nil,
  multiple_testing_correction: correction() | nil,
  options: map() | nil,
  tests: [test()]
}

test()

@type test() ::
  :ttest | :bootstrap | :anova | :mannwhitney | :wilcoxon | :kruskal | atom()