Interpreting correlations

Lab 4D

Directions: Follow along with the slides and answer the questions in red font in your journal.

Some background...

  • So far, we’ve learned about measuring the success of a model based on how close it’s predictions come to the actual observations.
  • The correlation coefficient is a tool that gives us a fairly good idea of how these predictions will turn out without having to make predictions on future observations.
  • For this lab, we will be using the movie data set to investigate the following questions:

Which variables are better predictors of a movie's audience_rating when the predictions are made using a line of best fit?

Correlation coefficients

  • The correlation coefficient describes the strength and direction of the linear trend.
  • It's only useful when the trend is linear and both variables are numeric.