Correlation coefficient (r) is a number that indicates the degree of linear relationship between two variables.
I. Scatter diagram (scattergram, scatterplot) is graphic representation of r. Each dot on scattergram represents two measures on one unit.
II. Reference values
A. 1 or +1 indicates perfect predictability and that both variables increase/decrease together (height in inches vs. cm. example).
B. -1 also indicates perfect predictability, but that one variable increase while the other decreases (height in inches vs. distance between head and ceiling example).
C. 0 indicates random relationship with no predictability.
D. All r's are between these points. This is only an ordinal scale.
-1_______________0_______________+1
E. However, there are some limitations.
1. The decimals cannot be converted into percentages, because it is only an ordinal scale.
2. Absolute value can never be greater than 1 (see formula).
3. Correlation does not necessarily mean causation
r = .3 indicates mild relationship with minimal predictability.
r = .95 indicates strong relationship with great predictability.
Explain the concept of correlation and how to interpret correlation coefficients of 0.3,0, and -0.95
1 answer
1 answer