Multiple Regression & Multiple Correlation — Previous Year Questions

📊 Topic Overview

Metric	Value
Papers Appeared	8 of 27 (29.6%)
Priority	🟡 MEDIUM
Section Distribution	Sec A: ~2
Most Repeated	"Multiple regression (short note)" / "Discuss multiple correlation. Explain partial correlation."

🎯 Must-Know Concepts

1. Multiple regression: predicting one criterion from two or more predictors
2. Multiple correlation coefficient (R): correlation between criterion and all predictors combined
3. R² = coefficient of multiple determination
4. Difference from simple (bivariate) regression
5. Assumptions of multiple regression

---

🔥 Most Repeated Question

Common phrasings:

• "Discuss multiple correlation. Explain partial correlation with suitable example."
• "Multiple regression (short note)."
• "Define regression. Differentiate between linear and multiple regression by citing example."
• "When do we use partial and multiple correlations?"

Answer (Section B or C — ~250 words):

Multiple Regression:

Multiple regression is an extension of simple linear regression in which two or more predictor variables (X₁, X₂, ..., Xk) are used to predict a single criterion variable (Y).

Equation: $\hat{Y} = a + b_1X_1 + b_2X_2 + b_3X_3 + \ldots + b_kX_k$

Where:

• a = regression constant (Y-intercept)
• b₁, b₂, ... = partial regression coefficients (slope for each predictor)
• Each b coefficient represents the change in Y for a one-unit change in that predictor, holding all other predictors constant

Example: Predicting academic performance (Y) from study time (X₁), intelligence (X₂), and motivation (X₃) simultaneously.

Multiple Correlation (R):

The multiple correlation coefficient (R) is the correlation between the observed values of Y and the predicted values of Y (Ŷ) obtained from the multiple regression equation. It represents the overall relationship between the criterion and all predictors combined.

Properties:

• R ranges from 0 to +1 (always positive, unlike Pearson's r)
• R² = coefficient of multiple determination — the proportion of variance in Y explained by all predictors together
• R ≥ r_YX for any single predictor (adding predictors never decreases R)

Simple vs Multiple Regression:

Aspect	Simple Regression	Multiple Regression
Predictors	1	2 or more
Equation	Ŷ = a + bX	Ŷ = a + b₁X₁ + b₂X₂ + ...
Correlation	Pearson's r	Multiple R
Variance explained	r²	R²
Complexity	Simple	More complex

Assumptions of Multiple Regression:

1. Linear relationship between each predictor and criterion
2. No multicollinearity (predictors should not be highly correlated with each other)
3. Normality of residuals
4. Homoscedasticity (equal variance of residuals)
5. Independence of observations

---

🧠 Memory Aid

Multiple R: Multiple predictors → Multiple R → R² tells us how much variance all predictors together explain in Y.

---

📚 Related Content

• Standardised Multiple Regression
• Semi-partial and Multiple Correlation
• Regression Analysis: Introduction

✅ Self-Assessment

1. Write the multiple regression equation for two predictors.
2. What does R² represent in multiple regression?
3. Why is R always positive while Pearson's r can be negative?
4. What is multicollinearity and why is it a problem?
5. How does multiple regression differ from simple regression?

💡 Exam Tips

• Section C (3 marks): Multiple regression = 2+ predictors → Ŷ = a + b₁X₁ + b₂X₂ → R = multiple correlation → R² = variance explained.
• Often asked alongside partial correlation — prepare both together.
• The comparison table (simple vs multiple regression) is a strong Section B answer.
• Know R² — it is one of the most frequently tested concepts within this topic.