LINER Conditions
Linear · Independent · Normal · Equal variance · Random
Checked mainly with a residual plot & residual histogram.
Hypotheses
H₀: β = 0 vs Hᵣ: β ≠ 0 (or >, <)
β = true population slope; b = sample slope estimate.
Degrees of Freedom
df = n − 2
Two parameters (intercept & slope) are estimated first.
Standard Error of Slope
SE_b = s / √Σ(x − x̄)²
s = residual standard deviation.
Confidence Interval
b ± t* · SE_b
t* comes from the t-distribution at df = n−2.
Test Statistic
t = (b − β₀) / SE_b
Measures how many SE_b's b is from β₀ (usually 0).