Linear Regression & Diagnostics
Ordinary Least Squares (OLS), ANOVA partition, collinearity diagnostics (VIF, Tolerance), and Durbin-Watson autocorrelation.
Verified vs SciPy (≤ 10⁻⁶)IBM SPSS Algorithms pp. 883–894
1. OLS Formulation & Normal Equations#
Linear regression solves the coefficient vector via QR/Cholesky decomposition of the normal equations:
StatCraft outputs full standardized coefficients , standard errors, -values, and two-tailed significance levels.
2. Collinearity Statistics (VIF & Tolerance)#
For each predictor , the tolerance and Variance Inflation Factor are computed to detect multicollinearity.
3. Residual Autocorrelation#
Durbin-Watson is computed from residual differences to verify independent errors:
Canonical SPSS Syntax Example
SPSS SYNTAX (.SPS)
| 1 | REGRESSION |
| 2 | /DEPENDENT salary |
| 3 | /METHOD=ENTER educ prevexp |
| 4 | /STATISTICS=COEFF R ANOVA COLLIN TOL ZPP. |