Correlation Analysis

INFO

statistical technique used to measure the strength and direction of the relationship between 2 or more variables

• widely used in business and data science to determine patterns and dependencies that can inform strategic decision-making
• range from -1 to +1
  • values closer to +1 indicate strong positive relationship
  • values closer to -1 indicate strong negative relationship
  • values around 0 suggest no significant relationship
• commonly used correlation metrics
  • Pearson’s correlation (measures linear relationships)
  • Spearman’s rank correlation (monotonic relationships)
  • Kendall’s Tau correlation (ordinal data)

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from scipy.stats import pearsonr
 
# Generate synthetic data
np.random.seed(42)
advertising_spend = np.random.randint(5000, 20000, 50)  # Advertising budget
revenue = advertising_spend * 3.5 + np.random.normal(0, 5000, 50)  # Revenue with some noise
 
# Create a DataFrame
data = pd.DataFrame({'Advertising Spend': advertising_spend, 'Revenue': revenue})
 
# Calculate Pearson Correlation
corr_coefficient, p_value = pearsonr(data['Advertising Spend'], data['Revenue'])
 
# Display correlation matrix
plt.figure(figsize=(6,4))
sns.heatmap(data.corr(), annot=True, cmap='coolwarm', fmt=".2f")
plt.title("Correlation Matrix")
plt.show()
 
# Print results
print(f"Pearson Correlation Coefficient: {corr_coefficient:.2f}")
print(f"P-Value: {p_value:.5f}")

• The Pearson correlation coefficient provides insight into the strength of the relationship between advertising spend and revenue
  • If output shows a correlation coefficient close to
    • +1: indicating a strong positive → higher advertising expenditures are generally associated with higher revenue
• p-value tests the statistical significance of the correlation
  • p-value < 0.05: statistically significant

IMPORTANT

correlation DOES NOT establish a causal link, other factors could also influence revenue

• strengths of correlation analysis
  • ability to quickly identify relationships between variables
    • valuable tool for exploratory analysis
  • provides quantitative insights
• Weakness of correlation analysis
  • inability to determine causation
    • high correlation between 2 variables $\ne$ one causes the other
    • hidden confounding variables could be influencing both
  • most effective for linear relationships
    • if the relationship is non-linear, standard correlation coefficients may provide misleading results
  • impact of outliers
    • can distort correlation values and lead to incorrect conclusions
  • To mitigate these issues, correlation analysis should be complemented with regression modeling and experimental design to establish causal relationships