This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Descriptive Statistics for Quantitative Data  I Notation • Suppose we have n experimental units and are interested in a single variables that is quantitative in nature. – Let xi denote the value of the variable for the i th experimental unit. • X = { x 1, x 2, . . . , xn } would be the collection of the observed values for the variable of interest for all n observations. – In other words, X is the dataset. Notation • Example 1: Suppose our dataset was A = {187.6, 125.3, 210.8, 157.3, 140.6}. Thus n = 5. • Could have x 1 = 187.6, x 2 = 125.3, x 3 = 210.8 x 4 = 157.3 and x 5 = 140.6. • Or could label x 1 = 140.6, x 2 = 125.3, x 3 = 157.3 Notation • In this instance (and throughout the course)the observation that we call x 1, the one we call x2 , and so on, is arbitrary. – It is merely just a convention to refer back to the observation. • In general it is not always arbitrary. An example would be if the observations occur in a sequence. – For instance, if the observations are recorded over time, the subscript would give Measures of Center Introduction • Question: Given the data { x 1, x 2, . . . , xn }, what would be a good guess as to what the value of a ‘typical’ observation is? • Answer: Measures of center provide an answer to the above question. • Three commonly reported measures of center are: the sample mode, sample median, and sample mean (sample Sample Mode • The sample mode is simply the observation that occurs the most frequently in the data set. • Example 2: Suppose the dataset is B = {2, 4, 6, 3, 2}. For this example the sample mode would be 2, since it occurs the most. • Problem – The sample mode may not be Sample Mode • Back to A = {187.6, 125.3, 210.8, 157.3, 140.6}. Here each observation occurs once and only once. • Example 3: Suppose the dataset is C = {2, 4, 4, 2, 6}. Here both 2 and 4 occur twice. • In both instances the sample mode is not unique. This happens more often than Sample Median • Intuitively, the sample median is the number that divides the data into two equal halves. However, this is not precise. A better definition is needed for the sample median. To do this, we will need to introduce another concept and notation, Order Statistics. Order Statistics • Given the data { x 1, x 2, . . . , xn }, define x ( i ) to be the ith smallest value. • So x (1) is the smallest value, x (2) the second smallest value, . . . , and x ( n ) the last smallest (i.e. the largest) value. • If two or more observations are tied, than necessarily the same number of order statistics must have the same value. Order Statistics • Back to Example 1: The data was x 1 = 187.6, x 2 = 125.3, x 3 = 210.8, x 4 = 157.3 and x 5 = 140.6....
View
Full Document
 Spring '09
 POPEL
 Statistics, sample median

Click to edit the document details