Regression lines: Line of best fit minimizes sum of squared errors. Two possible lines: (1) Y on X: ŷ = a + bx (predict y from x); (2) X on Y: x̂ = c + dy (predict x from y). Rarely same line unless perfect correlation. Intersection point: (x̄, ȳ) (always passes through). Properties: Both lines pass through mean point. Angle between lines smaller if correlation stronger. Example: r = 0.9, sy=20, sx=10. Line Y on X: b = 0.9×2 = 1.8. Line X on Y: d = 0.9×0.5 = 0.45. For r=0 (perpendicular); r=1 (same line). Solving: Calculate each regression line. Plot both. Compare predictions. Exam tip: Understand why two lines exist. Know when to use each. Practice: Dual regression problems.