Chapter 3: Numerical Descriptions of Data

Objectives
 Calculate measures of center (mean, median, mode)
 Calculate measures of dispersion (range, variance, standard deviation, IQR)
 Calculate percentiles, quartiles, and five number summary
 Discuss about the measures affected/unaffected by outliers
 Construct a boxplot using five number summary

Two ways to summarize observed data

 Graphically, e.g., pie charts, histograms, dot plots, etc. (Chapter 2)
 Numerically, e.g., mean, median, variance, etc. (Chapter 3)

Measures of Center
 Mean
 Median
 Mode
Measures of Dispersion
 Range
 Variance
 Standard deviation
 Interquartile range (IQR)
Measures of Position
 Percentile
 Quartile
 z-score

Measures of Center
  Numerical values used to locate the “middle” of a set of data, or where the
data is clustered.
 The term average is often associated with all measures of center.
 Be careful when you see the word average in things like newspapers…you
won’t know which average is meant!

Mean: The mean is the sum of all observations divided by the total number of
observations.
Sample Mean: Population Mean:
n N

∑ xi ∑ xi
x= i=1 µ= i=1
n N

where n is the size of the sample, N is the size of the population, Σ means to
sum/add

Example 3.1: The following data represents a sample of the number of accidents
per month in the last 8 years at a particularly dangerous intersection. Find the
mean number of accidents: 8, 9, 3, 5, 2, 6, 4, 5.

Weighted Average
To calculate a weighted average—like grades in this class, you multiply the weight
by the average in that category and add them up.

Example 3.2: Students often like to calculate their final grade. Suppose the
assignment grades of a student in our STAT 200 are given as follows. Note, the
weights for each assignment category is also given in our syllabus. Calculate the
student’s final grade point average (GPA).
Category Weight Student’s average in
each category
Quiz 15% 70
Homework 20% 92
Exam I 15% 85
Exam II 15% 75
Exam III 15% 72
Final Exam 20% 82

Median: The value of the data that occupies the middle position when the data
are ranked in order according to size

Steps to Calculate Median

1. List the values in ascending order (lowest to highest).
2. Find the middle-most score. That’s the median.
3. If two scores are in the middle, take average of those two middle values.

Example 3.3: Find the median income from five different households:
$135456, $25500, $32456, $54365, $37668.

Example 3.4: Find the median income from six different households:
$135456, $25500, $32456, $54365, $37668, $34500.

Example 3.6: CO2 Pollution of the Eight Largest Nations: 0.2, 0.7, 1.1, 1.2, 1.8, 2.3,
9.8, 19.7 (19.7 seems to be an outlier). A dot plot is presented below.

a. Find the mean and median of the data set:

Mean =
Median =
a. Now, remove the outlier and find the mean and median again.
 Mean =
Median =

Question: Which measure of center, mean or median, is more influenced by

outliers (extreme observations)?

Mode: The mode is the observation(s) of a data set that occurs/occur most
frequently.

Note that
 there is no mode if each observation occurs only once or if all observations
occur equally.
 a data set is unimodal if there is a unique value that occurs most frequently.
 a data set is biomodal if two observations occur equally most frequently.
 a data set is multimodal if more than two observations occur equally most
frequently.
Example 3.5: Determine the mode of each set of data below:
a. 2, 4, 5, 7
b. 2, 2, 5, 5, 9, 9
c. 3, 4, 5, 5, 5, 6, 7, 7, 7, 7, 12
d. 3, 4, 5, 5, 5, 6, 7, 7, 7, 12
e. 1, 2, 2, 3, 3, 4, 4
f. 5, 5, 5, 6, 6, 7, 7, 7, 8, 9, 6

Outliers
 Although the mean, the median, and the mode each describe a typical entry
of a data set, there are advantages and disadvantages of using each,
especially when the data set contains outliers.
 An outlier is a data entry that is far removed from the other entries in the
data set. (There are mathematical methods to determine if a particular item
is an outlier, but in our class we will “eyeball” it.)

When to use what

 Use the mode when the data are categorical and values can fit into only one
class, such as hair color or political affiliation.
 Use the median when you have extreme scores or outliers and you don’t
want to distort the average. Income is often an example where there are
extremes.
 Use the mean when you have data that do not have extremes and are not
categorical.
Example 3.7: Look at the following examples to determine which value of central
tendency to use. Why?
(i) Mean (ii) Median (iii)Mode

a. Grades in a stat 200 class at the end of the semester:

AAABBBBBBBBBBBBCCCCCCCCCCCCCCCCCCCCCDDDDDFF

b. Weights of freshman offensive football players at a local high school (in lbs.)
110, 145, 150, 151, 156, 160, 162, 175, 175, 177, 293

c. Calorie content of a typical meal of lasagna, bread, salad and dessert at 6 local
restaurants: 2500, 2700, 3000, 3320, 3500, 3700

Calculator instructions
Put your data into a list:
Stat>Edit>
1-var-stats:
Stat>calc>1:1-var stats(list)

Example 3.8: Find the mean, median, and mode of the data below using your
calculator: 7,6,3,3,1
Measures of Dispersion
 Measures of central tendency alone cannot completely characterize a set of
data. Two very different data sets may have similar measures of central
tendency.
 Measures of variation are used to describe the spread, or variability, of a
distribution
 Common measures of dispersion: Range, variance, and standard deviation

Range: The difference in value between the highest-valued (H) and the lowest-
valued (L) pieces of data:
Range = H – L
Ex. 7, 6, 3, 3, 1 Ex. 3, 4, 4, 5, 4 Ex. 4, 4, 4, 4, 4

Sample variance & sample standard deviation

Sample Variance: The sample variance, s2, is the mean of the squared deviations,
calculated using n - 1 as the divisor, where n is the sample size:

Sample Standard Deviation: The sample standard deviation, s, is the positive

square root of the sample variance:
Population variance & population standard deviation
Population Variance: The population variance, σ2, is the mean of the squared
deviations, calculated using N as the divisor, where N is the population size:

Population Standard Deviation: The population standard deviation, σ, is the

positive square root of the population variance:

***Another measure of dispersion is the interquartile range (IQR) we learn later.

Example 3.10: Calculated variance and standard by hand for the following sample
data: 7, 6, 3, 3, 1.
Data x x−x (x−x )
2

Total
Example 3.9: Characterize the following by either statistic or parameter.
a. Sample variance:
b. Sample standard deviation:
c. Population variance:
d. Population standard deviation:

Odds and ends

 Variance and standard deviation can NEVER be negative.
 The standard deviation is the measure of spread of the data that is easiest to
understand because it has the same units as the original data.
 Standard deviation is a measure of the typical amount an entry deviates
from the mean. The more the entries are spread out, the greater the
standard deviation.
 Standard deviation is very sensitive to outliers (just like the mean) so it could
be large because of one unusually large or small value.
 If s=0 there is no variability in the data. Ex. 4, 4, 4, 4, 4 has s = 0.

Calculator instructions
To find mean, sample standard deviation and population standard deviation,
sample size, min, first quartile, median, third quartile, and max very quickly:
Enter data into a list
Stat>Calc>1-Var Stats>
List #>

Be careful to use the correct standard deviation: Population or sample.

Example 3.11: Find the Standard Deviation and the Variance for the following
sample data using your calculator: 7, 6, 3, 3, 1.

Example 3.12: Each histogram represents a set of data with a sample mean of five.
Clearly they are very different sets of data.
• Which one has the largest standard deviation?
• Which one has the least?

Percentiles, Quartiles, and Boxplots

 Sometimes we want to specify a position of a certain data entry within a
data set. Quartiles and percentiles are two of the most popular measures of
position
Quartiles
Q1 = First quartile (25th percentile); 25% of the data fall below Q1
Q2 = Second quartile (50th percentile, median); 50% of the data fall below Q2
Q3 = Third quartile (75th percentile); 75% of the data fall below Q3

Steps to calculate Q1 and Q3 (approximation):

 Sort the data into ascending order (small to large), split the data set into two
halves by median, and exclude the median.
 Calculate the median of the lower half; this is Q1.
 Calculate the median of the upper half; this is Q3.

Five number summary: The five-number summary consists of

Interquartile range (IQR)

 IQR = Q3 – Q1

Note: IQR is resistant to outliers. So, in the presence of outliers, IQR is a better
measure than standard deviation or range to compare the variability among
different data sets.
Box-and-Whisker plot:
Box-and-Whisker plot, also called boxplot, can be used to summary quantitative
data. This is one of the most common data summarization techniques.
Constructing a boxplot requires to calculate the five number summary. See the
example below.

Example 3.13: Calculate the (i) five-number summary, (ii) IQR, and (iii) construct a
boxplot for the following data: 2, 3, 3, 5, 7, 9, 10, 12, 15.

Example 3.14: Calculate the (i) five-number summary, (ii) IQR, and (iii) construct a
boxplot for the following data: 2, 3, 3, 5, 7, 9, 10, 12, 15, 20.
Calculator instructions for 5 number summary
Stat > Calc > 1:1-Var Stats (L#)

Calculator instructions for boxplots

Input data into List(s)
2nd > Stat Plot > 1:plot1…
Turn it on by highlighting On
Type: Middle plot, bottom row (box plot)
Xlist: L1
Freq: 1
Zoom >9:Zoomstat > Trace >>>>

***To enter multiple boxplots, repeat by doing Stat Plot >2 and 3 if necessary (You
can turn on 3 plots at once and see all 3 boxplots at one time.)

Percentiles: Values of the variable that divide a set of ranked data into 100 equal
subsets; each set of data has 99 percentiles. The kth percentile, Pk, is a value such
that k% of the data are smaller in value than Pk and at most (100 - k)% of the data
is larger.
 A student’s SAT score is 82-th percentile. What does it mean?

Notes:
 The 1st quartile and the 25th percentile are the same: Q1 = P25
 The median, the 2nd quartile, and the 50th percentile are
all the same: ~x=Q2 =P 50.
Example 3.15:
 What percent of data is below P23?

 What percent of data is above the 45th percentile?

Standard Scores, or z-score

Standard score (z-score) measures relative position of data values from their
mean. More specifically, a z-sore tells how many standard deviations an
observation is from the mean.
The z-score corresponding to the observation x from a population with mean μ
and standard deviation σ is given by

Example 3.16: If the mean GRE score is 650 with standard deviation of 100, what is
the z-score for a student with (i) 770, (ii) 500, (iii) 650. Interpret the z-scores.

Example 3.17: Calculate the z-score for an observation x = 20 where the mean and
variance of the corresponding population are 12 and 16. Interpret the z-score.
Empirical Rule: For (ONLY) bell shaped distribution
 about 68% of observations fall within 1 standard deviation of the mean
 about 95% of observations fall within 2 standard deviations of the mean
 about 100% of observations fall within 3 standard deviations of the mean

Example 3.18: The average college student produces 640 pounds of solid waste each year
(Source: [Link] Assume the distribution of waste per
college student is bell-shaped with a mean of 640 pounds and a standard deviation of 60
pounds. Sketch the distribution using the empirical rule.
a. 95% of college students produce between ____________ and ____________ pounds of
solid waste each year.

b. 68% of college students produce between ____________ and ____________ pounds of

solid waste each year.

c. About what percent of students produce more than 700 pounds of solid waste each
year? Shade on bell-shaped curve:
d. About what percent of students produce less than 520 pounds?