【正文】
tliers if 8 + = Continuous Data Summary ? A robust statistic is resistant to extreme values. ? Robust measures of center and spread should be used for distributions that are nonsymmetric, multimodal, or contain outliers to ensure quality of the results. Tradit ion a l R ob ust C e nt e r x (M e a n ) ~x (M e d ian ) Sprea d S (Std Dev ia tion ) Va r ian ce R (R a n g e ) IQR Discussion Scenarios Scenario 1 ? When are the mean and median similar? ? The mean and median will be similar when the distribution is symmetric. ? Why it is important to take outlier effects into account? ? Outliers can bias estimates of the mean, standard deviation, and variance. Discussion Scenarios (Continued) Scenario 2 ? Which is a more representative measure of center for the following variables? – Housing prices. ? Median – Salaries of all Intel employees. ? Median – Heights of students in this statistics class. ? Mean – Heights of students in this statistics class with an NBA basketball star in attendance. ? Median – Student grades over four years of school. ? Mean Discussion Scenarios (Continued) Scenario 3 ? Calculate the mean and median of the following values: 5, 9, 2, 4, and 10. ? Mean = 6 ? Median = 5 ? Calculate the mean and median of the following values: 5, 9, 2, 4, and 100. ? Mean = 24 ? Median = 5 Median did not change. Practice Items 1. Define the statistics that are used to measure center – give an example of when each should be used. ? Mean is the arithmetic average of a distribution. Median is the center point of a distribution. Mean should be used in most cases unless the distribution is highly skewed. 2. What is the ‘Sigma Rule’ and why is it important? ? The ‘Sigma Rule’ states that m and s can be used to describe the entire distribution. It is important because it allows us to easily estimate the sprea