Question 1
1 markKai calculated Spearman's rank correlation coefficient for two sets of data. He got the value 1.2
Explain how you know that Kai has made a mistake.
Model answer
Spearman's rank correlation coefficient always lies between -1 and +1. Kai's value, 1.2, is greater than 1, so he must have made a mistake.
Explanation
In the formula, 6Σd² ÷ n(n² - 1) can never be negative, because Σd² is a sum of squares, so r_s can never be more than 1. The value 1 means the two rankings are exactly the same and -1 means they are exactly opposite. A value outside -1 to 1 is a sign of an arithmetic error, such as a wrong Σd².
Mark scheme
-
1 mark
B1 for Spearman's rank correlation coefficient must be between -1 and 1 (it cannot be more than 1)
oe, eg 'the largest possible value is 1'. 'It should be negative' B0. 'It is too big' with no reference to 1 B0. 'Correlation coefficients cannot be decimals' B0.