How To Calculate Correlation In Math Example

Calculate Pearson, Spearman, or Kendall correlation between two datasets

Correlation measures the statistical relationship between two continuous variables. Understanding how to calculate correlation is fundamental in statistics, research, and data analysis. This comprehensive guide explains correlation coefficients, calculation methods, and practical applications with real-world examples.

Key Takeaways

• Correlation quantifies the strength and direction of a relationship between variables
• Pearson’s r measures linear relationships (-1 to +1)
• Spearman’s ρ and Kendall’s τ assess monotonic relationships
• Statistical significance determines if the relationship is meaningful
• Visualization with scatter plots helps interpret correlation

1. Understanding Correlation Basics

1.1 What is Correlation?

Correlation is a statistical measure that describes the degree to which two variables move in relation to each other. It answers three key questions:

1. Direction: Do variables increase together (positive) or move in opposite directions (negative)?
2. Strength: How closely are the variables related?
3. Form: Is the relationship linear or non-linear?

Correlation Coefficient (r) = Covariance(X,Y) / (σX × σY)

1.2 Types of Correlation Coefficients

Coefficient	Measures	Range	When to Use
Pearson’s r	Linear relationships	-1 to +1	Normally distributed data, linear trends
Spearman’s ρ	Monotonic relationships	-1 to +1	Ordinal data or non-linear trends
Kendall’s τ	Ordinal associations	-1 to +1	Small datasets or many tied ranks

1.3 Correlation vs. Causation

A common statistical fallacy is confusing correlation with causation. Remember:

• Correlation indicates variables move together
• Causation means one variable directly affects another
• Example: Ice cream sales and drowning incidents are correlated (both increase in summer), but one doesn’t cause the other

2. Calculating Pearson Correlation Coefficient

2.1 Pearson’s r Formula

The Pearson correlation coefficient (r) measures linear correlation between two variables X and Y:

r = [n(ΣXY) - (ΣX)(ΣY)] / √{[nΣX² - (ΣX)²][nΣY² - (ΣY)²]}

Where:

• n = number of data pairs
• ΣXY = sum of products of paired scores
• ΣX = sum of X scores
• ΣY = sum of Y scores
• ΣX² = sum of squared X scores
• ΣY² = sum of squared Y scores

2.2 Step-by-Step Calculation Example

Let’s calculate Pearson’s r for this dataset (Study Hours vs. Exam Scores):

Student	Study Hours (X)	Exam Score (Y)	X²	Y²	XY
1	5	82	25	6724	410
2	3	75	9	5625	225
3	7	91	49	8281	637
4	2	65	4	4225	130
5	6	88	36	7744	528
ΣX = 23		ΣY = 401	ΣX² = 123	ΣY² = 32599	ΣXY = 1930

Applying the formula with n = 5:

1. Numerator: (5 × 1930) – (23 × 401) = 9650 – 9223 = 427
2. Denominator X: (5 × 123) – (23)² = 615 – 529 = 86
3. Denominator Y: (5 × 32599) – (401)² = 162995 – 160801 = 2194
4. Final denominator: √(86 × 2194) = √188684 ≈ 434.38
5. r = 427 / 434.38 ≈ 0.983

2.3 Interpreting Pearson’s r Values

r Value Range	Interpretation
0.90 to 1.00	Very high positive correlation
0.70 to 0.90	High positive correlation
0.50 to 0.70	Moderate positive correlation
0.30 to 0.50	Low positive correlation
0.00 to 0.30	Negligible or no correlation
-0.30 to 0.00	Low negative correlation
-0.50 to -0.30	Moderate negative correlation
-0.70 to -0.50	High negative correlation
-1.00 to -0.70	Very high negative correlation

3. Spearman Rank Correlation

3.1 When to Use Spearman’s ρ

Use Spearman’s rank correlation when:

• Data is ordinal (ranked)
• Relationship appears non-linear
• Data contains outliers
• Assumptions of Pearson’s r aren’t met

3.2 Calculation Steps

1. Rank each value in both variables (1 = smallest)
2. Calculate differences between ranks (d)
3. Square the differences (d²)
4. Apply formula: ρ = 1 – [6Σd² / n(n²-1)]

3.3 Example Calculation

Using the same study hours and exam scores data:

Student	Study Hours (X)	Rank X	Exam Score (Y)	Rank Y	d	d²
1	5	3	82	3	0	0
2	3	2	75	2	0	0
3	7	5	91	5	0	0
4	2	1	65	1	0	0
5	6	4	88	4	0	0
Σd² = 0

ρ = 1 – [6 × 0 / 5(25-1)] = 1 – 0 = 1.00 (perfect correlation)

4. Kendall’s Tau Correlation

4.1 Understanding Kendall’s τ

Kendall’s tau is particularly useful for:

• Small datasets (n < 30)
• Data with many tied ranks
• Ordinal data analysis

4.2 Calculation Method

Kendall’s tau compares the number of concordant (aligned) and discordant (misaligned) pairs:

τ = (C - D) / √[(C + D + T)(C + D + U)]

Where:

• C = number of concordant pairs
• D = number of discordant pairs
• T = number of ties in X only
• U = number of ties in Y only

5. Testing Statistical Significance

5.1 Why Test Significance?

Significance testing determines whether an observed correlation:

• Is likely due to real relationship
• Or could occur by random chance

5.2 Hypothesis Testing

For correlation tests:

• Null Hypothesis (H₀): ρ = 0 (no correlation)
• Alternative Hypothesis (H₁): ρ ≠ 0 (correlation exists)

5.3 Critical Values Table

For Pearson’s r with n=5 at α=0.05 (two-tailed), the critical value is 0.878. Our calculated r=0.983 exceeds this, so we reject H₀.

Sample Size (n)	Critical r (α=0.05, two-tailed)	Critical r (α=0.01, two-tailed)
5	0.878	0.959
10	0.632	0.765
20	0.444	0.561
30	0.361	0.463
50	0.279	0.361

6. Practical Applications of Correlation

6.1 Business and Economics

• Stock market analysis (how different assets move together)
• Demand forecasting (relationship between price and sales)
• Risk management (portfolio diversification)

6.2 Healthcare and Medicine

• Drug dosage vs. effectiveness studies
• Lifestyle factors and disease risk correlations
• Genetic marker associations with conditions

6.3 Education Research

• Study habits and academic performance
• Teaching methods and learning outcomes
• Socioeconomic status and educational attainment

7. Common Mistakes to Avoid

7.1 Misinterpreting Correlation Strength

• Even “strong” correlations (r=0.7) leave 51% of variance unexplained (r²=0.49)
• Context matters – r=0.3 might be significant in social sciences but weak in physics

7.2 Ignoring Non-Linear Relationships

Always visualize data with scatter plots. Pearson’s r only detects linear relationships:

7.3 Violating Assumptions

Pearson’s r requires:

• Both variables are continuous
• Linear relationship
• Normally distributed variables
• No significant outliers
• Homoscedasticity (equal variance across values)

8. Advanced Topics

8.1 Partial Correlation

Measures relationship between two variables while controlling for others:

rxy.z = (rxy - rxzryz) / √[(1-rxz²)(1-ryz²)]

8.2 Multiple Correlation

R (multiple correlation coefficient) measures how well multiple predictors explain a dependent variable:

R = √(1 - [SSres/SStot])

8.3 Correlation Matrices

Used in multivariate analysis to show all pairwise correlations in a dataset:

	Height	Weight	Age	Income
Height	1.00	0.68	0.12	0.05
Weight	0.68	1.00	0.08	0.15
Age	0.12	0.08	1.00	0.42
Income	0.05	0.15	0.42	1.00

9. Learning Resources

For further study, consult these authoritative sources:

• NIST/Sematech e-Handbook of Statistical Methods – Comprehensive statistical reference from the National Institute of Standards and Technology
• UC Berkeley Statistics Department – Academic resources on correlation and regression analysis
• CDC’s Principles of Epidemiology – Practical applications of correlation in public health research

Pro Tip

Always complement correlation analysis with:

• Scatter plots to visualize relationships
• Regression analysis to model the relationship
• Effect size measures to quantify practical significance
• Confidence intervals for precision estimation