A researcher is planning to perform a Pearson's correlation on scale data but notices several extreme outliers that appear to skew the distribution.What is the best alternative method to analyze the data?
A) Transform the data to ordinal rankings and perform a Spearman correlation.
B) Transform the data to ratio variables and perform a Spearman correlation.
C) Use the scale data to run a Spearman correlation.
D) Throw out the outliers and run a Pearson correlation on the scale data.

Question

Quizplus · Accepted Answer

Transforming the data to ordinal rankings and performing a Spearman correlation is a better alternative than using the scale data for a Pearson correlation when there are extreme outliers that skew the distribution. The Spearman correlation is based on ranking the data, so extreme outliers will have less influence on the correlation coefficient.

A Researcher Is Planning to Perform a Pearson's Correlation on Scale