Linear & Non-linear Relationship — Previous Year Questions
📊 Topic Overview
| Metric | Value |
|---|---|
| Papers Appeared | 7 of 27 (25.9%) |
| Priority | 🟡 MEDIUM |
| Section Distribution | Sec A: ~3 |
| Most Repeated | Usually paired with a Spearman's Rho computation: "Explain linear and non-linear relationship. Find out the degree of relationship using Spearman's Rho." |
🎯 Must-Know Concepts
- Linear relationship: constant rate of change; scatter plot fits a straight line
- Non-linear relationship: variable rate; scatter plot follows a curve
- Positive and negative linear relationships
- Pearson's r only valid for linear relationships
- Outliers, curvilinearity, and their effect on correlation
🔥 Most Repeated Question
Common phrasings:
- "Explain linear and non-linear relationship with suitable diagrams."
- "Define outliers. Compute Spearman's Rho for the following data." (linear/non-linear is the preamble)
- "Describe linear and non-linear relationship with suitable examples."
- "Linear and Non-linear Relationship (short note)."
Answer (Section A conceptual component — ~200 words):
Linear Relationship:
A linear relationship between two variables X and Y exists when the change in Y is proportional to the change in X at a constant rate. When plotted on a scatter diagram, the data points cluster around a straight line.
Types:
- Positive linear relationship: As X increases, Y increases proportionally. The regression line slopes upward (e.g., study hours and exam scores).
- Negative linear relationship: As X increases, Y decreases proportionally. The regression line slopes downward (e.g., anxiety level and performance quality).
Pearson's product moment correlation coefficient (r) and linear regression are appropriate measures for linear relationships.
Non-linear Relationship:
A non-linear (curvilinear) relationship exists when the change in Y is not proportional to the change in X. The scatter plot follows a curved pattern rather than a straight line.
Example: The Yerkes-Dodson law describes an inverted U-shaped relationship between arousal (X) and performance (Y). Performance increases with moderate arousal but decreases at both very low and very high arousal levels. Pearson's r would underestimate this relationship (giving r near 0).
Implication: If Pearson's r ≈ 0, it does not necessarily mean no relationship — it may indicate a non-linear relationship. Visual inspection of the scatter plot is essential.
Outliers:
Outliers are extreme data points that deviate markedly from the rest of the distribution. They can artificially inflate or deflate Pearson's r. Spearman's rho is less sensitive to outliers because it uses ranks.
🧠 Memory Aid
- Linear = straight line on scatter plot → Pearson's r is appropriate
- Non-linear = curved pattern → Pearson's r underestimates; use eta (η) or Spearman's rho
📚 Related Content
✅ Self-Assessment
- What is the difference between a linear and a non-linear relationship?
- Can Pearson's r = 0 if there is a strong non-linear relationship? Explain.
- Give one psychology example of a non-linear relationship.
- How does an outlier affect Pearson's r?
- Draw a scatter plot showing a positive linear relationship and a curvilinear relationship.
💡 Exam Tips
- Section C (3 marks): Linear = straight line, constant rate → Non-linear = curved → Pearson's r only valid for linear → Example (Yerkes-Dodson).
- Questions almost always come with a Spearman computation — the theory component is brief (1 paragraph each for linear and non-linear).
- Always draw labelled scatter diagrams — they earn marks.
- Mention outliers and their effect on Pearson's r (leads naturally to why Spearman's rho is preferred in such cases).