Suppose that ten new smart phone models were evaluated by two consumer electronics magazines (Popular Electronics and Electronics Now) from 1 (best) to 10 (worst) as shown below. Why is the Spearman's rho more appropriate than the correlation coefficient for these data?

A) The relationship is nonlinear.
B) The data are ordinal.
C) The data are not paired.
D) There is one extreme outlier.
E) Both outliers and bends in the data make the correlation coefficient difficult to interpret.

Question

Suppose that ten new smart phone models were evaluated by two consumer electronics magazines (Popular Electronics and Electronics Now) from 1 (best) to 10 (worst) as shown below. Why is the Spearman's rho more appropriate than the correlation coefficient for these data?

A) The relationship is nonlinear. 
B) The data are ordinal. 
C) The data are not paired. 
D) There is one extreme outlier. 
E) Both outliers and bends in the data make the correlation coefficient difficult to interpret.

Quizplus · Accepted Answer

The data are ordinal, meaning that the rankings have a natural order but the distance between each ranking is not necessarily equal. Spearman's rho is a better measure of the relationship between the rankings because it accounts for the ordinal nature of the data. The correlation coefficient assumes that the data are interval or ratio, which is not the case here.

Suppose That Ten New Smart Phone Models Were Evaluated by Two