A random sample of statistics students was selected. The heights (inches) and the weights (pounds) of the students were collected and are displayed in the figure below. There is a student whose measurements were outliers in the data set. What do you expect will happen to the correlation if that student's measurements are removed?
A) The correlation will be greater than (further from zero) it was.
B) The correlation will be less than (closer to zero) it was.
C) The correlation will remain the same as it was.
D) There is not enough information to determine how the correlation might change.

Question

A random sample of statistics students was selected. The heights (inches) and the weights (pounds) of the students were collected and are displayed in the figure below.  There is a student whose measurements were outliers in the data set. What do you expect will happen to the correlation if that student's measurements are removed?
A) The correlation will be greater than (further from zero) it was. 
B) The correlation will be less than (closer to zero) it was. 
C) The correlation will remain the same as it was. 
D) There is not enough information to determine how the correlation might change.

Quizplus · Accepted Answer

Outliers can have a strong effect on the correlation coefficient, especially if they have extreme values in both variables. Removing an outlier is likely to decrease the correlation coefficient, making it closer to zero.

A Random Sample of Statistics Students Was Selected