(i) What is the meaning of positive and negative correlation? Explain with suitable example.
(ii) Define regression. What are coefficients of regression and what are their properties? Give examples.
15 marks · directive: Define
The core of a good answer
Correlation is positive when two variables move in the same direction and negative when they move in opposite directions; regression estimates the average value of one variable from another through a line whose slopes, the regression coefficients b(yx) and b(xy), have the same sign as r and satisfy r² = b(yx) × b(xy).
An outline written by upsc.help, not an official UPSC answer; every fact checked by a separate review. How we work
What the question asks
Part (i) (5 marks) wants the meaning of positive and negative correlation with examples (and the range of r); part (ii) (10 marks) wants a definition of regression, the two regression coefficients with formulas, their properties, and an illustration.
Answer outline
Introduction
Correlation measures the degree and direction of linear association between two variables; regression goes further and expresses one variable as a function of the other for estimation and prediction.
(i) Positive and negative correlation
- Positive: both variables increase or decrease together; r lies between 0 and +1. Example: advertising expenditure and sales; height and weight.
- Negative: one increases as the other decreases; r lies between -1 and 0. Example: price and quantity demanded; machine age and output.
- r = +1 or -1 is perfect correlation; r = 0 means no linear correlation (not necessarily no relation).
- Karl Pearson's coefficient r = Σ(x - x̄)(y - ȳ) / √[Σ(x - x̄)² Σ(y - ȳ)²].
(ii) Regression and its coefficients
- Regression is the estimation of the average value of a dependent variable from given values of an independent variable using a functional (usually linear) relation.
- Regression of Y on X: Y - ȳ = b(yx)(X - x̄), with b(yx) = r σy/σx = Σ(x - x̄)(y - ȳ) / Σ(x - x̄)².
- Regression of X on Y: X - x̄ = b(xy)(Y - ȳ), with b(xy) = r σx/σy = Σ(x - x̄)(y - ȳ) / Σ(y - ȳ)².
- A regression coefficient gives the change in the dependent variable for a unit change in the independent variable.
Properties of regression coefficients
- r = ±√[b(yx) × b(xy)]: r is the geometric mean of the two coefficients.
- Both coefficients and r have the same sign.
- If one coefficient exceeds 1 (in absolute value), the other must be less than 1, since their product r² ≤ 1.
- Their arithmetic mean is at least r in absolute value: |b(yx) + b(xy)|/2 ≥ |r| (AM ≥ GM, as both have the same sign).
- They are independent of change of origin but not of scale.
- The two regression lines intersect at (x̄, ȳ); they coincide when r = ±1 and are perpendicular when r = 0.
Worked example
- X: 1, 2, 3, 4, 5; Y: 2, 4, 5, 4, 5. x̄ = 3, ȳ = 4.
- Σ(x - x̄)(y - ȳ) = 6; Σ(x - x̄)² = 10; Σ(y - ȳ)² = 6.
- b(yx) = 6/10 = 0.6; Y = 2.2 + 0.6X.
- b(xy) = 6/6 = 1.0; X = Y - 1.
- r = √(0.6 × 1.0) = 0.775, positive like both coefficients; their mean 0.8 ≥ 0.775.
Conclusion
Correlation shows whether and how strongly variables move together; regression coefficients turn that association into an estimating equation used in sales forecasting, cost estimation and demand analysis.
Examples and data to cite
- Francis Galton coined 'regression' in the nineteenth century from his studies of heights of parents and children (Regression towards mediocrity in hereditary stature, 1886).
- Management use: estimating sales from advertising spend, or maintenance cost from machine age.
Facts to remember
- Range of r: -1 ≤ r ≤ +1.
- r² (coefficient of determination) is the share of variance in Y explained by X.
- b(yx) × b(xy) = r².
- Correlation does not imply causation; spurious correlation can arise from a common third factor.