0% found this document useful (0 votes)
32 views13 pages

Understanding Mean and Median in Data Analysis

Module 5 focuses on measures of center, specifically mean and median, and their application in analyzing data distributions. It includes exercises on calculating these measures, understanding their implications in skewed distributions, and recognizing when to use each measure. The module emphasizes the importance of understanding the shape of data distributions and how outliers can affect mean and median values.

Uploaded by

gabeaguzman2006
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
32 views13 pages

Understanding Mean and Median in Data Analysis

Module 5 focuses on measures of center, specifically mean and median, and their application in analyzing data distributions. It includes exercises on calculating these measures, understanding their implications in skewed distributions, and recognizing when to use each measure. The module emphasizes the importance of understanding the shape of data distributions and how outliers can affect mean and median values.

Uploaded by

gabeaguzman2006
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Module 5 – Measures of Center

Module 5.1 – A Feel for Measures of Center

Learning Goal: For the distribution of a quantitative variable, describe the overall pattern
(shape, center, and spread) and striking deviations from the pattern.
Specific Learning Objectives:

 Find the mean and median from different representations of data.


 Develop number sense with mean and median by creating different data sets with a
given mean or median.

N = 11
Set-up/Warm-up: The dot plot gives quiz scores for a small class.

a) What is the mean? Show your work or explain how you got your answer.
Average

5788999991010 = 93

93/11 = 8.455 = 8.455


Mean :

b) What is the median? Show your work or explain how you got your answer.
Middle #
N+1 11 + 1
Location of median 12/2 = 6
------- -------
2 2

6th data value is the median


6th value is 9 so median is 9
c) Which measure (the mean or the median) is the better way to represent typical
performance on this quiz? Why?

Median because distribution is skewed left.

Always use median if graph is skewed

Page 31
Exercises:

1) Here are two sets of exam scores, one for a class that has 4 students and one for a class
that has 15 students.

Class A: 80, 90, 90, 100


Class B: 60, 65, 65, 70, 70, 70, 75, 75, 80, 80, 80, 80, 80, 85, 100

a) Without doing any calculations, which class do you think will have a larger mean?
Why?

Class A because the scores are higher with none of them going below 80
while Class B has multiple 60's and 70's.

b) Now calculate the mean for each class. Which is larger? Why does this make sense?

Class A = 90
Class B = 75.67

2) For this data, is the mean wrist measurement for men (larger than, smaller than, or
about the same as) the mean wrist measurement for women? (Obviously, you can’t
calculate the means, so jot down notes about how you thought about this.)

11.25

9.75

Males have larger wrists according to the data

Page 32
3) For this data, is the mean pelvic diameter for men (larger than, smaller than, or about
the same as) the mean pelvic diameter for women? (Obviously, you can’t calculate the
means, so jot down notes about how you thought about this.)

Males have slightly higher pelvic diameters

4) This table gives quiz scores for a different class. Scores Number of
Students
a) What is the mean? Show your work or explain how 5 1
you got your answer. 6 3
5666777778889 = 91 7 5
8 3
91/13 = 7 9 1
n = 13

b) What is the median? Show your work or explain how you got your answer.

Median is 7th data value


N1
----- 13 + 1 14/2 = 7
2 ------- Median is 7
2

c) Which measure (the mean or the median) is the better way to represent typical
performance on this quiz? Why?

Both work for this data set, but should use the mean
because the distribution is symmetrical

Page 33
5) For this problem, use the digits 0, 1, 2, 3, 4, 5, 6, 7, 8, 9.

a) List five digits that have a median of 7 and a mean of 7 (repeats allowed). Find a
different set of 5 digits that work.

5, 6, 7, 8, 9

6, 6, 7, 8, 8

b) List five digits that have a median of 7 and a mean that is less than 7 (repeats
allowed.) Give the mean of your 5 digits. Find a different set of 5 digits that work.

1, 2, 7, 8, 9 5.4

3, 4, 7, 8, 9 6.2

c) List five digits that have a median of 7 and a mean that is more than 7 (repeats
allowed.) Give the mean of your 5 digits. Find a different set of 5 digits that work.

5, 6, 7, 9, 9 = mean = 7.2
6, 6, 7, 8, 9 = mean = 7.2

6) For this problem, use the digits 0, 1, 2, 3, 4, 5, 6, 7, 8, 9. Construct a data set where
neither the mean nor the median is a reasonable “typical” value.

0, 0, 0, 9, 9, 9

median = 4.5
09
----- 4.5
2

Page 34
7) Imagine that you have a bag filled with 9 numbers. The mean and the median of the
numbers in the bag are both 6.

a) You draw a number out of the bag. It is a 4. You replace it with a 1. Does the mean of
the numbers in the bag get bigger, smaller, or stay the same? What about the
median? Jot down some notes to explain how you figured this out.
Original Data
3, 4, 5, 6, 6, 6, 7, 8, 9
N=9 If I replace the 4 with a 1 the mean gets smaller because
54 it would be 51/9 5.667 instead of 54/9 6 , however
91 --- Mean = 6
the median stays the same since the 5th value is still 6.
---- 5th value 9
2
Median is 6
b) You draw a number out of the bag. It is an 8. You replace it with 8 ones. Does the
mean of the numbers in the bag get bigger, smaller, or stay the same? What about
the median? Jot down some notes to explain how you figured this out.

3, 4, 5, 6, 6, 6, 7, 1, 1, 1, 1, 1, 1, 1, 1, 9

54 The mean does get smaller at 3.375 and


N = 16 Mean = 3.375
--- the median gets smaller as well at 1.
161 16
------ 8.5th value
2 Median is 1
8) Even though statisticians call the mean and median measure of “center”, it might be less
confusing to think of the mean and the median as ways to summarize a distribution with
a single number. In other words, the mean and the median are ways to represent a
“typical “measurement in a data set. When do you think is it better to use the median
and when it is it better to use the mean to summarize a distribution?

I think it is better to use the median when a distribution is skewed to a


certain direction (left or right). However I think it is better to use the mean
when a distribution is symmetrical.

Page 35
Module 5.2 – The Mean as a Balancing Point
Learning Goal: For the distribution of a quantitative variable, describe the overall pattern
(shape, center, and spread) and striking deviations from the pattern.
Specific learning Objective: Discover why the mean is called a balancing point.

1) Betty Sue is struggling in her math class but her effort is steadily improving. To date, her
quiz scores are 0, 43, 56, 66, 68, 74, 83, and 90 (each quiz is worth 100 points). What is her
mean score?
043566668748390480

480
----- 60
8
a) Student A has the same mean score as Betty Sue, but his quiz scores are all the same
score. Make a dotplot of Student A’s 8 quiz scores and mark the mean.
60 x 8 480
480
---- 60
8

b) Student B also has the same mean score as Betty Sue. She has 6 quizzes with a score of
60 and one quiz with a score of 70. What did she score on the 8 th quiz? Make a dotplot
of Student B’s 8 quiz scores and mark the mean. Show how you figured this out.

5060606060606070 480

480
60
----
8

Page 36
c) Student C also has the same mean score as Betty Sue. She has 5 quizzes with a score of
60 and two quizzes with a score of 70. What did she score on the 8 th quiz? Make a
dotplot of Student C’s 8 quiz scores and mark the mean. Show how you figured this out.

560270X 300140X 440 + X = 480


60 ------------- 60
---------------- X = 40
8 8

d) Student D also has the same mean score as Betty Sue. Is it possible for Student D to
have a 100 on a quiz? What is the maximum number of 100-point quiz scores possible?
How do you know?

2041004480
60 Max number of 100 score is 4.
8

2) Draw two dotplots of 8 quiz scores such that


 Both have the same mean as Betty Sue’s (mark the mean) 60
 One dotplot has very little spread
 One dot plot has a lot of spread

Little
Spread

Large
Spread

Page 37
3) Write Betty Sue’s mean quiz score below. For each of her quiz scores, record the signed
distance from the mean by calculating Score – Mean. A score above the mean will have
positive distance while a score below the mean will have negative distance.
60
Betty Sues Mean quiz score: ________

-60 -17 -4 6 8 14 23 30

Now add up the signed distances from the mean. What did you get?
-60174681423300

4) For one of the distributions you created in problem #2, record the quiz scores in the table
below. For each quiz score, record its signed distance from the mean.
60
Mean Quiz Score (which is the same as Betty Sue’s): ________

55 55 60 60 60 60 65 65

-5 -5 0 0 0 0 5 5

Now add up the signed distances from the mean. What did you get?
-55550

5) What do you notice when you add the signed distances from the mean? Why do you think
this happens?

6) For the dotplot of quiz scores below, find the mean. If the student takes another quiz and it
doesn’t change his mean score, what did he score on this quiz? Explain why this makes
sense using signed distances from the mean.

Page 38
7) The student with the distribution of quiz scores below takes two more quizzes. One score is
above his mean and the other is below his mean but his mean does not change. What are
some possible pairs of quiz scores for this student? Show three possibilities using the
dotplots below.

Mean = 80

75 85

8) Why do you think that statisticians call the mean the “balancing point” of a distribution?

I think they call the mean the "balancing point"


because it's the exact average or center of all
the numbers so if it were on a scale the mean
would be the perfect point of balance.

Page 39
Module 5.3 – Shape and Measures of Center

Learning Goal: For the distribution of a quantitative variable, describe the overall pattern
(shape, center, and spread) and striking deviations from the pattern.

Specific Learning Objective:

 Relate measures of center to the shape of the distribution.


 Choose the appropriate measure for different contexts.

Warm-up/Set-up: Up to now, we’ve covered two types of Measures of Center: the mean and
the median. What we’ll see in this module is that though both values can be used to describe
your distribution, both of these values are susceptible to outliers. That is, having an outlier will
drag both the mean and the median in a particular direction. So, there are instances where one
measure of center is better than the other.
1) Here is a dot plot of Hilda’s quiz scores in her math class.
N13

a) What is the shape of this distribution?

The shape is symmetrical.

b) Determine Hilda’s median quiz score. Calculate Hilda’s mean quiz score.

I approximate that the median score is 8.

667788888991010104
mean = 8
13

Page 40
2) Oops, we made a mistake recording two of Hilda’s scores. The two scores recorded as 6
should have been 0. Here’s the revised dot plot of her quiz scores.
N13

a) What is the shape of this distribution?

The shape is both skewed left and roughly symmetrical

b) Find her revised median quiz score. Is it less than, more than, or about the same as her
previous median quiz score? Why does this make sense?

131 7th value It should still be the same median score because the 7th value is
2 still 8 just like the last score.

c) Find her revised mean quiz score. Is it less than, more than, or about the same as her
previous mean quiz score? Why does this make sense?

00778888899101092 mean = 7.076


13

The mean is less than her pervious mean.

d) Which measure of center (mean or median) would best represent her typical
performance on the quizzes? Explain.

In this case the median would best represent her typical


performance because majority of her scores were around the
median score of 8 and the only reason the mean was thrown off
was because of the two outlier "0" scores.

Page 41
3) Write a few sentences that explain how the shape of the distribution might influence
whether the mean is larger than, smaller than or about the same as the median.

Most of the time when a distribution is skewed the mean is going to be larger or smaller then the
median. Whether it's larger or smaller depends on the way it's skewed and the type of outliers that
are present. For example if the distribution is skewed left and has outliers in the lower end (like 0's
and 10's) then the mean will be smaller than the median, however if the distrubion is skewed right
and the outliers are in the higher end (like 90's and 100's) then the mean will be larger than the
median.

4) For each distribution give an estimate for the median. Then say whether the mean is
probably greater than, less than, or about equal to the median. Jot down some notes to
explain how you figured this out.

50% ----- Median ----- 50%

5) Which of the following distributions is likely to have a mean that is smaller than the
median? Jot down some notes to explain how you figured this out.

a) Scores on the a Module Checkpoint, where almost all of the grades were A’s and B’s
(scores of 80 to 100), with one student making a zero.

b) Repeated weighing of “one-pound” bags of peanut M&M’s using a digital bathroom


scale that reports weights to the nearest tenth of a pound.

c) Salaries of NBA basketball players (This distribution is strongly right skewed by the few
superstars who make much more money than the rest of the players.)

Page 42
6) According to the U.S. Bureau of Census, the median family income in the United States in
2008-2012 was $53,046. Why do you think the median, rather than the mean, was reported
by the U.S. Bureau of Census?

I think they used the median rather than the mean because of the outliers of people
with too much income or too little income that would throw off the balance.

7) Summarize what you have learned about how to choose between the mean and median as
the best measure of center based on the shape of the distribution.

I think we should choose to use mean when the shape is mostly symmetrical
and there are no outliers that can skew the data, instead I think that's where
the median should be used, when the distribution shape is skewed (left or right).

Page 43

Common questions

Powered by AI

The U.S. Bureau of Census likely chooses to report the median family income instead of the mean due to income outliers that could skew the mean. Extremely high or low incomes would affect the mean more dramatically, making it less representative of the typical family income, whereas the median provides a more stable central value in such cases .

The mean is considered the 'balancing point' of a distribution because it is the arithmetic average that balances the sum of the deviations (positive and negative) from itself. This means that the total distance of values below the mean is equal to the total distance of the values above it, creating a conceptual point of balance .

A dataset where neither the mean nor the median represents a typical value could be constructed with values like [0, 0, 0, 9, 9, 9]. Here, the mean and median might suggest a central tendency of 4.5, which does not match any value within the set or concentrate around it. This discrepancy shows the limitations when significant gaps exist between groups of numbers .

Changing individual data points affects the mean and median differently: the mean will change more significantly with any addition or subtraction of values due to its nature as the arithmetic average. In contrast, the median remains unaffected unless the changes cross the midpoint of the data when ordered, particularly if the changes do not cross the data value that determines the median position .

The median wrist measurement for men is typically larger than for women due to biological differences where men generally have larger bones and muscles. This suggestion leverages general physiological trends without requiring specific data calculations, as it reflects aggregate observations in human anatomy .

If a number is replaced in the dataset, the mean is directly affected since it's calculated as the sum of all values divided by the count. For instance, replacing a 4 with a 1 will decrease the mean to 5.67 as the sum decreases by the difference between these numbers. The median, however, remains unchanged if the replaced number is not at the median position such as the central value when ordered, so the median would still be 6 unless a median value itself is altered .

Outliers influence the mean significantly because they shift the average of all values, particularly if they are extreme. Conversely, the median is more resistant to such changes, as it depends on the middle value of the ordered dataset, which only shifts if an outlier alters the dataset configuration around the midpoint .

The median is more appropriate to use when a distribution is skewed left or right because it is less affected by extreme values or outliers. Since the mean can be significantly influenced by such outliers, it may not represent the typical value for a skewed distribution as accurately as the median does .

A symmetrical distribution with a mean of 8 might include scores like [6, 7, 8, 8, 9, 9, 10]. A skewed distribution keeping the same mean could have scores like [2, 6, 8, 9, 9, 9, 10, 15], where additional extreme values pull the spread while maintaining the central tendency. This creates variability in skewness while retaining the same mean .

Yes, Student D can have a maximum of four 100-point quiz scores and still maintain the same mean score of 60 as Betty Sue. This is possible if the remaining scores are adjusted to maintain the total sum of 480 across eight quizzes. Specifically, having four scores of 100 contributes 400 points, allowing the remaining four quizzes to sum to 80, e.g., they could be all 20s, ensuring the mean remains at 60 .

You might also like