Why a Correlation Coefficient of r=0.9 Can Be Practically 'Irrelevant'—The Gap Between Numerical Strength and Practical Significance

While looking at data, one sometimes realizes something: even when the correlation coefficient reaches r=0.9, it may actually ...

Pitfalls of Statistical Analysis: Your Decision to Determine 'No Correlation' Is Likely 80% Wrong

You, looking at data and deciding, 'This has zero correlation.' That conclusion might be a statistical illusion.The Pearson correlation coefficient is beautiful. It can express the linear relationship ...