2.2.11 An Introduction to Linear Regression - Video 6: Correlation and Multicollinearity

TL;DR

Correlation measures the linear relationship between two variables and is represented by a number between -1 and +1.

Transcript

In the previous video, we observed that Age and FrancePopulation are highly correlated. But what is correlation? Correlation measures the linear relationship between two variables and is a number between -1 and +1. A correlation of +1 means a perfect positive linear relationship. A correlation of -1 means a perfect negative linear relationship. In ... Read More

Key Insights

• 📶 Correlation measures the linear relationship between two variables, indicating the strength and direction of the relationship.
• 💯 A correlation of +1 or -1 represents a perfect linear relationship, while a correlation of 0 means no linear relationship.
• 💻 Correlation can be visually examined through scatter plots but is more accurately computed using statistical methods.
• 🥺 High correlation between independent variables can lead to multicollinearity issues in linear regression models.
• 🏃 Removing insignificant variables can help address multicollinearity, but caution should be exercised to avoid removing important variables.
• ❓ Coefficients in linear regression models should be interpreted in the presence of other variables.
• 🍉 Correlations above 0.7 or below -0.7 are generally considered to be cause for concern in terms of multicollinearity.

Questions & Answers

Q: What does correlation measure?

Correlation measures the strength and direction of the linear relationship between two variables.

Q: How is correlation represented?

Correlation is represented by a number between -1 and +1, with -1 indicating a perfect negative linear relationship and +1 indicating a perfect positive linear relationship.

Q: What does a correlation of 0 mean?

A correlation of 0 means that there is no linear relationship between the two variables.

Q: How are correlations computed in R?

Correlations can be computed in R using the "cor" function, which takes the names of the two variables as input and returns the correlation value.

Summary & Key Takeaways

• Correlation measures the linear relationship between two variables, ranging from -1 to +1, with 0 indicating no linear relationship.

• Various examples are provided to demonstrate different levels of correlation between variables.

• The R programming language is used to compute correlations between variables in a dataset.