Power Analysis

hypotestx.power.sample_size.n_ttest_two_sample(effect_size: float, alpha: float = 0.05, power: float = 0.8, alternative: str = 'two-sided') → int

Minimum n per group (balanced) for an independent-samples t-test.

Parameters:

• effect_size (Cohen's d)

• alpha (Type I error rate)

• power (desired power)

• alternative ('two-sided', 'greater', or 'less')

Returns:

int

Return type:

minimum n per group; total N = 2 * returned value

hypotestx.power.analysis.power_ttest_two_sample(effect_size: float, n1: int, n2: int | None = None, alpha: float = 0.05, alternative: str = 'two-sided') → float

Post-hoc power for an independent-samples t-test.

Parameters:

• effect_size (Cohen's d ((mean1 - mean2) / pooled_sigma))

• n1 (first group size)

• n2 (second group size (default = n1, i.e. balanced design))

• alpha (significance level)

• alternative ('two-sided', 'greater', or 'less')

Returns:

float

Return type:

estimated statistical power (0..1)

Notes

NCP = d * sqrt(n1*n2/(n1+n2)).

Functions Reference

n_ttest_two_sample

Calculate the required sample size per group to detect a given effect size at the desired power level for an independent two-sample t-test.

import hypotestx as hx

# Medium effect, standard alpha, 80% power
n = hx.n_ttest_two_sample(effect_size=0.5, alpha=0.05, power=0.8)
print(f"Required n per group: {n}")   # 64

# Small effect, 90% power
n = hx.n_ttest_two_sample(effect_size=0.2, alpha=0.05, power=0.9)

power_ttest_two_sample

Compute the achieved power for a two-sample t-test given the observed effect size and sample sizes.

pow_result = hx.power_ttest_two_sample(
    effect_size=0.4,
    n1=30,
    n2=30,
    alpha=0.05,
)
print(f"Achieved power: {pow_result.power:.2f}")