Script for Correlation Audio

Modified: 2024-05-29 9:39 AM CDST

Here is the script for the audio file on correlation coefficients

• Correlational methods look for relationships between two or more variables.
  • The do not, however, assess, cause and effect. They only assess the degree of relationship.
  • Correlation coefficients range from +1.00 to -1.00.
  • The sign of a correlation indicates its direction, either positive or negative.
  • Look at the cheerleader figure. The two variables are indicated by the pompoms, each labelled A or B.
  • When both pompoms are each high or each low, that indicates a positive correlation.
  • When one pompom is high and the other low or vice versa that indicates a negative correlation.
  • So, when A and B run in the same direction, high or low, that is a positive correlation. When A and B run in opposite directions that is a negative correlation.
• Correlation coefficients, then, can be used to predict B when A is known or to predict A when B is known.
  • Prediction confidence increases the higher the numerical value of the correlation is.
• Here are two examples, the first one is a negative correlation.
  • You just bought a bicycle and are showing it to me. I try to pick it up, it’s heavy.
    • So now I can predict you bought a cheap bicycle. A, the weight, is high, so I infer that B, the price, is low.
  • A year later you bought another bicycle, I go to pick it up and it’s light.
    • I say, this is an expensive bike. Why? Again, A, the weight, is low, so I infer the price is high. BTW, some bicycles cost over $5000. Guess what? They better be light.
• What about a positive correlation? Look at p. 34. A is study time and B is test grade. So, if study time is high then grade should be high. If study time is low, then grade should be low.
• Once again, remember, correlations only describe the relationship between variables they cannot, by themselves, assess that one variable caused the other.

Back to Chapter 1 and 2 Lectures