If Measurement Error in Our Variables Is High When Attempting

Multiple Choice

If measurement error in our variables is high when attempting to calculate correlation coefficients, which of the following three approaches is not a good way to deal with this?

A) Find a different measuring method that yields more precise measurements.
B) Measure each individual multiple times and use the average value.
C) Remeasure individuals with the most extreme values to eliminate outliers.
D) Use repeat measurements to calculate a corrected correlation, r*.