WEEK 1
FIVE CHARACTERISTICS OF DISTRIBUTION
. Mean
. Standard deviation
. Skewness
. Kurtosis
. Z-scores and outliers
The Z-score of an observation shows how many standard deviations this
observation lies below (negative score) or above (positive score) the sample
mean
– Identifies potential outliers
PERCENTILES AND BOX PLOT
Location of pth percentile of sample Lp = (n+1)(p/100)
Lower (upper) hinge = the midpoint of the lower (upper) half of the data set
Turkey’s original method includes the median in both halves of the data set /
field excludes the median
5
Upper fence = Q3 + (1.5*IQR)
4
Lower fence Q1 - (1.5*IQR)
3
– Values lying outside of the fences are outliers
2
1
Whisker = the lines where observations in each tail that are not outliers
Kurtosis can be seen by comparing the length of each of the two tails with
FIVE CHARACTERISTICS OF DISTRIBUTION
. Mean
. Standard deviation
. Skewness
. Kurtosis
. Z-scores and outliers
The Z-score of an observation shows how many standard deviations this
observation lies below (negative score) or above (positive score) the sample
mean
– Identifies potential outliers
PERCENTILES AND BOX PLOT
Location of pth percentile of sample Lp = (n+1)(p/100)
Lower (upper) hinge = the midpoint of the lower (upper) half of the data set
Turkey’s original method includes the median in both halves of the data set /
field excludes the median
5
Upper fence = Q3 + (1.5*IQR)
4
Lower fence Q1 - (1.5*IQR)
3
– Values lying outside of the fences are outliers
2
1
Whisker = the lines where observations in each tail that are not outliers
Kurtosis can be seen by comparing the length of each of the two tails with
