Outliers
freq dist, histogram, stats ***
47 50 48 51 52 55 51 53 54 52
n=___ mean=______ SD=______ Outliers?:______
Is the mean close to the median:___ and look at them in the boxplot.
Change the first datum 47 to 30.
n=___ mean=______ SD=______ Outliers?:______
Some change of the mean. How is this new mean in relation to the IQR:_________ (boxplot)
Percentage change of the SD:______% ==(new-old)/old *100%
Outliers can have a big effect on the SD because the (x-mean) squared
has an outsized effect. Very big effect when the data set is small.
Some data distros
Copy and paste the "n=100 normally distributed, u=50, s=5" following line of numbers.
n=____ mean=______ SD=______ Outliers?:______
Change the first datum 40 to 0.
n=____ mean=______ SD=______ Outliers?:______
Much change of the mean?:____
Percentage change of the SD:______% ==(new-old)/old *100%
Outlier still has an effect on the SD of this larger data set but not as
much as on a small data set.
The more outliers, the more the effect on the SD.