This preview shows page 1. Sign up to view the full content.
Unformatted text preview: e Correct Charts & the Area Principle 0 Mean Salaries at a Major University, 2002  2005 20000 40000 60000 Correct Charts & the Area Principle 02 03 04 05 Graphs for Quantitative Variables
Histogram
To make a histogram:
1. Divide the range of the data set
into equal intervals.
2. For each interval draw a bar
with the base covering the
interval and the height
proportional to the the count of
observations that fall into the
interval. #.Facebook Wallposts Summary Statistics: Measures of Center
v༇ Summary Statistics
q༇ Numerical Descriptions of Distributions.
q༇ Measures of center: Ø། MODE: ü༏ Categorical variable: category with highest frequency. ü༏ Numerical variable: location of a major peak of the distribution. Ø། MEDIAN: Middle Value . ü༏ Which value is the 50th Percentile. Ø། MEAN: AVERAGE MEAN: More detail
(and the Σ notation)
v༇ Example: Consider data for a variable is 6, 9, 8, 3, 3, 1. 6 + 9 + 8 + 3 + 3 +1
Mean of the variable =
=5
6
v༇ For a variable x with n observed values x1 , x2 ,..., xn
the mean of x is given by:
n x1 + x2 + + xn
x=
=
n ∑x i i =1 n Median: the 50th Percentile
Arrange data in order.
Median Md = 50th percentile = “middle observation”
[if number of observations is even, average the middle two.] E.g. for data 1, 3, 3, 6, 8
Md = 3 E.g. for data 1, 3, 3, 6, 8, 9 Md = (3 + 6) / 2 = 4.5 “Robust” or “resistant”
v༇ Robust = insensitive to a few extreme observations
(imagine a typo of adding several zeros to a number) v༇ Which is more robust: mean or median ?
Compare 1, 3, 3, 6, 8
to
1, 3, 3, 6, 8000000
v༇ Comparison of 3 different measures of center
MEASURES Applicable for Variable Uniq...
View Full
Document
 Fall '07
 Lv

Click to edit the document details