Professional Documents
Culture Documents
1 - 2 ProbStat Lectures Summary Statistics 1 - 2
1 - 2 ProbStat Lectures Summary Statistics 1 - 2
1
Measures of Central Tendency
• Sample Mean
• Median
• Mode
2
Example
• Consider a sample of the hours spent streaming videos in a week:
Observation No. 1 2 3 4 5
Hours 5 3 7 43 7
3
Sample Mean:
4
Median
• Middle term of ordered set.
• Divides the data into two “equal” parts.
5
Mode
• Most frequent value.
• 3, 5, 7, 7, 43 → Mode = 7
• 3, 3, 5, 5, 7 → Mode = 3, 5 (Bimodal)
• 3, 4, 5, 6 → No mode.
6
• Why multiple measures of central tendency?
• Advantages and disadvantages to each.
• Mean is easy to calculate, but sensitive to outliers.
• Median more robust to outliers.
Observation No. 1 2 3 4 5
Hours 5 3 7 43 7
• Mean = 13
• Median = 7
7
Measures of Variability
• Sample Variance
• Sample Standard Deviation
• Percentiles
• Quartiles
• Range
• Interquartile Range
8
Sample Variance, s2
• n observations
• ith observation value Xi
• Sample mean
9
Formulas for s2
• Definition Formula:
• Computational Formula:
• Alternate Formula:
10
Standard Deviation, s
• Standard deviation has the same units as and
11
Example – Height Measurements
• Suppose that the heights of a sample of people are measured. The
heights in inches are:
• 66, 74, 64, 69, 61, 59, 70, 72
• Compute the sample mean, variance, and standard deviation
12
Example – Height Measurements
• Suppose that the heights of a sample of people are measured. The heights in
inches are:
• 66, 74, 64, 69, 61, 59, 70, 72
• Compute the sample mean, variance, and standard deviation
• Solution:
• in.
• in.2
•
∑
• in.2
• in.
13
13
Example cont.
• What if the measurements are converted to centimeters?
• Let X be the sample measurement in inches, and let Y be the sample
measurement in centimeters.
14
Example cont.
• What if the measurements are converted to centimeters?
• Let X be the sample measurement in inches, and let Y be the sample
measurement in centimeters.
•
•
•
•
15
Example cont.
• What if everyone grew 3 extra cm?
16
Example cont.
• What if everyone grew 3 extra cm?
•
•
•
17
Other Measures of Variability
• Range = Max – Min
• Interquartile Range (IQR) = Q3 – Q1, where Q1 and Q3 are the first and
third quartile
18
Percentiles
• pth percentile is the value such that p% of the sample is below and (1-
p)% is above
• Steps:
• 1: Order the sample from smallest value to largest
• 2: Compute:
• 𝑃𝑜𝑖𝑛𝑡𝑒𝑟 = 𝑛+1
• 3: If Pointer is an integer, then it points to the pth percentile in the ordered
sample list. If Pointer is not an integer, then average the sample values that
Pointer is between
19
Percentile Example
• Example:
20
Percentile Example
• Example:
• Solution:
•
• Pointer points between 5 and 6 in ordered sample list.
• 30th percentile is
21
Quartiles
• Q1 = 25th percentile
• Q2 = 50th percentile (same as median)
• Q3 = 75th percentile
22
Quartile Example
• Compute: Q1, Q2, Q3, and IQR for the following sample:
• 2, 4, 5, 8, 9
23
Quartile Example
• Compute: Q1, Q2, Q3, and IQR for the following sample:
• 2, 4, 5, 8, 9
• Solution
• Median = Q2 = 5
• Q1: ; Pointer between 1 and 2 in sample list;
; Q1 = 3
• Q3: 4.5; Pointer between 4 and 5 in sample list;
; Q3 = 8.5
• IQR = 8.5 – 3 = 5.5
24