answersLogoWhite

0

Outliers are observations that are unusually large or unusually small. There is no universally agreed definition but values smaller than Q1 - 1.5*IQR or larger than Q3 + 1.5IQR are normally considered outliers. Q1 and Q3 are the lower and upper quartiles and Q3-Q1 is the inter quartile range, IQR.

Outliers distort the mean but cannot affect the median. If it distorts the median, then most of the data are rubbish and the data set should be examined thoroughly.

Outliers will distort measures of dispersion, and higher moments, such as the variance, standard deviation, skewness, kurtosis etc but again, will not affect the IQR except in very extreme conditions.

User Avatar

Wiki User

12y ago

Still curious? Ask our experts.

Chat with our AI personalities

SteveSteve
Knowledge is a journey, you know? We'll get there.
Chat with Steve
ViviVivi
Your ride-or-die bestie who's seen you through every high and low.
Chat with Vivi
JordanJordan
Looking for a career mentor? I've seen my fair share of shake-ups.
Chat with Jordan

Add your answer:

Earn +20 pts
Q: What are outliers and how do they affect data?
Write your answer...
Submit
Still have questions?
magnify glass
imp