Extreme values in statistics most of the times can be troublesome. If a data point is so
Question:
Extreme values in statistics most of the times can be troublesome. If a data point is so extreme compared to the rest of the data (i.e. unusually high or unusually low), then the first thing to do as a researcher is to double-check and see if there is no mal-function of the equipment in use, or there is no recording error, or there is no unusual thing happening . ..that doesn't reflect the norms of the distribution of the data. If it does, we need to see how extreme an observation is to be considered as an outlier. 1) Briefly discuss how to detect outliers under different scenarios....assuming symmetry then with the absence of symmetry. What impact , if there is any, do they have on the model?
Elementary Statistics
ISBN: 978-0538733502
11th edition
Authors: Robert R. Johnson, Patricia J. Kuby