A regression model fitted to 100 observations has r=0.82r=0.82r=0.82. After removing one point recorded with the wrong unit, the correlation becomes r=0.61r=0.61r=0.61.
Explain why the original value of r r\,r was misleading.
State two checks that should be made before deleting any apparent outlier.
Explain why the corrected correlation still does not demonstrate causation.