True μ
0.0
True σ
2.0
N (per experiment)
Resample
β = 1/√(2(N−1)) =
—
abs. error ±σβ:
—
σ̂ range [σ(1−β), σ(1+β)]:
—
Exact P(|frac| ≤ β) [χ²]:
—
Monte-Carlo P:
—
mean bias:
—