Chapter 5: Understanding and Comparing Distributions
What two things should we never do with outliers?
1. Silently leave an outlier in place and proceed as if nothing were unusual. 2. Drop an outlier from the analysis without comment just because its unusual.
List the six major components of a box plot.
1. horizontal line at median 2. horizontal line at Q3 3. horizontal line at Q1 4. upper fence/whisker 5. lower fence/whisker 6. outliers 7. far outliers
Primarily, what are logs used for in statistics?
-to make data behave better
How are outliers displayed within a box plot?
-using a circle
How are far outliers displayed within a box polt?
-using an asterisk
When does a boxplot reveal skewness?
-when the median of the box is not centered between the quartiles -when the whiskers of a boxplot are not the same length
How do you calculate the lower fence that should be displayed within a box plot?
Lower Fence = Q1 - 1.5*IQR
How do you calculate the upper fence that should be displayed within a box plot?
Upper Fence = Q3 + 1.5*IQR
What is a timeplot?
-a graph that displays values against time
What is an outlier?
-any point more than 1.5 IQR and less than 3.0 IQR from either end of the box in a boxplot
What is a far outlier?
-any point more than 3.0 IQR from either end of the box in a boxplot
What are boxplots particularly good at?
-pointing out outliers
What can variables that are skewed to the right often benefit from?
-re-expression by square roots, logs, or reciprocals
What is the heigh of a box within a boxplot equal to?
-the IQR
