Numeracy

Introduction to Statistics
Statistics is a branch of mathematics that is concerned with the planning, collection, organisation, analysis,
reporting of data and the interpretation of results. The aim of this module is to give you an introduction to
statistics. This module gives you an overview of different methods of displaying and organising data,
calculating measures of central tendency, calculating measures of spread and comparing like data.
Correlation and Regression using Excel will inform the understanding of linear relationships.

Collecting Data
When data are collected from every member of the group, a census is held. The group in this instance is
called the population.

When only selected members of the population contribute to the data, this is referred to as a sample of the
population and a survey takes place. Mostly, it is impractical and too expensive to obtain data from a
population so a sample is selected. Information from a sample is often used to predict information about the
population. This is the process used to predict the outcomes of elections.

The selection of a sampling method is very important. For example it may be important to collect data from
different geographical locations, different ages or different socioeconomic backgrounds. For quality control
in industry, systematic sampling, taking (say) every tenth item to check quality could be used. It takes
careful planning to conduct investigations with little or no bias. Bias is an unfair preference towards one
group which may lead to a distortion of the statistical results.

The researcher must also decide on sample size. If the sample size is too small, then the validity of the
finding could be in doubt. If the sample size is too large, the cost of obtaining the data may be prohibitively
high. It is possible to obtain an appropriate sample size using statistical processes that considers the accuracy
of results required with the number to survey to give that accuracy.

Statistical data can be of different types, the type of the data may determine the statistical processes that can
be undertaken. Data can be classified as either Categorical or Numerical. Categorical data are usually in word
form. Numerical data are usually in number form, however, some data in number form such as postcodes
could be considered Categorical data as no statistical analysis would be performed; for example; the average
postcode.

There are two types of numerical data: discrete and continuous. Most of this module is about numerical
data.

Discrete data are numerical data where the values are in set amounts, for example, the number of people in a
classroom. There could be 25, 26, 27 people in a classroom but not any amounts in between such as 25.72
people.
Page 1
[last edited July 2024]
Student Learning Zone +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone
Continuous data are numerical data where the values can be any value, for example, the height of people. The
height of people can be measured, in theory, to any precision. The only limitation is the limitation of the
measuring instrument, but in theory the measurement can be to any precision.

Let’s consider some examples:

(a) Temperature – this is continuous because it can be measured to any precision.

(b) Number of students at a lecture – this is discrete because only a whole number of students could
attend.

(c) The number of children in a family – this is discrete because only a whole number of children is
possible.

(d) Age – this is continuous as age can be measured to any precision. For convenience, age in years may
be recorded – this is still continuous because age can be measured to any precision. The decision of
discrete vs continuous should be based on what is possible rather than what is chosen for convenience.

(e) Weight – this is continuous as weight can be measured to any precision.

(f) Shoe size – this is discrete because shoe size has set values of half sizes and none in between.

Page 2
 Numeracy

Module contents
Introduction Note: The tables in this section have not been
• Organising Data - Tables formatted for insertion in an assessment or
other publication. For table formatting requirements
• Graphs please refer to the style guide for your discipline
• Measures of Central Tendancy (e.g. APA7 or Harvard styles), and the guidance of
your Unit Assessor.
• Methods of Spread
• Comparing Like Data
• Correlation and Regression
Answers to activity questions
Outcomes
• Display data appropriately using charts and graphs.
• Organise data using tables.
• Calculate descriptive statistics for sets of data.
• Calculate correlation coefficient and equations of regression lines using Excel.

Check your skills

This module covers the following concepts, if you can successfully answer these questions, you do not need
to do this module. Check your answers from the answer section at the end of the module.

Data
1. For the data (right) about the Average Daily Hours of Location Latitude Av. Daily
Sunshine, calculate the mean, mode and median. °S Hours of
Sunshine
2. For the same data, calculate the range, interquartile Darwin 12.42 6.9
range and standard deviation. Brisbane 27.48 7.4
Perth 31.93 8.8
3. For the same data, use a 5 number summary to draw a Sydney 33.86 6.8
box and whisker plot. Adelaide 34.93 7.0
Canberra 35.3 7.2
4. Use Excel to do a scatterplot of Hours of Sunshine vs Melbourne 37.81 6
Latitude. Latitude is the independent variable Hobart 42.89 5.9

5. Use Excel to determine the correlation and the regression Macquarie 54.5 2.3
equation. Island

Page 3
[last edited July 2024]
Student Learning Zone +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone
 Numeracy

Topic 1: Organising Data - Tables

The quantity of data collected will determine how it is organised.

If the heights of 25 students are collected, then no table will be required. These 25 individual values can be
analysed without the need to organise the data.

If the grade point average of all the students at this university were to be analysed, then the data would need
to the organised into an appropriate table before any statistical analysis could take place. The structure of the
table will vary depending on the type of data (discrete or continuous) and the variation between lowest and
highest values. With spreadsheet computer programs such as Excel, the need to organise data using the
traditional table structures is not as important as in the past.

The first table to be considered is a Frequency Distribution Table. To assist in the development of ideas in
this module, some key examples will be used.

Ungrouped Frequency Distribution Tables

The data below are the numbers of matches per box in 50 boxes, these are discrete data. This number of values
is too many to have as individual values so organising the values into a table will assist when analysing the
data. The average number of matches per box is given as 50. The number of matches per box typically
varies between 46 and 55 matches. The raw data are given below:

49 50 50 49 50 48 50 53 48 53
51 47 54 48 52 50 47 55 49 55
52 50 51 51 50 50 53 49 52 48
54 49 46 50 50 49 50 50 50 54
46 51 51 47 48 50 52 50 51 50

When the data are initially entered into a table, tally marks are used to record each piece of data. Tally marks
are small vertical lines. After four tally marks, the fifth is put through the four to make grouping of 5.

When recording the occurrence of each value, data should be systematically entered. It is easy to make mistakes
when tallying, so be systematic and check your totals twice. Watch the video below to see how to do this.

Video ‘Entering Data into a frequency distribution table’

Page 4
Student Learning Zone +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone [last edited July 2024]
After the data are entered into the table they will have the appearance below. The Frequency is the number
of occurrences of that value or how frequently it occurs.

Number of Matches Tally Frequency

46 ΙΙ 2
47 ΙΙΙ 3
48 ΙΙΙΙ 5

49 ΙΙΙΙ Ι 6

50 ΙΙΙΙ ΙΙΙΙ ΙΙΙΙ Ι 16

51 ΙΙΙΙ Ι 6
52 ΙΙΙΙ 4
53 ΙΙΙ 3
54 ΙΙΙ 3
55 ΙΙ 2
Total 50

After this initial table is constructed, extra columns are added to the table to help with the summary data
required.

The Tally is not usually repeated.

Commonly added columns are Relative Frequency, %Relative Frequency and Cumulative Frequency.

The Relative Frequency is the proportion of the data that has that value. This can be expressed as a decimal
or fraction. It is calculated by taking the frequency for each score and dividing by the total number of
scores. Find a total for this column.

The %Relative Frequency is the Relative Frequency made into a percentage. This is achieved by
multiplying the Relative Frequency by 100. Find a total for this column.

The Cumulative Frequency is a running total of the frequency. The cumulative frequency is found on any
line of the table by adding the frequency for that line to the total of frequencies for the previous lines. Do not
obtain a total for the cumulative frequencies as it is already a (running) total. The cumulative frequency
column is used extensively for finding the median and the quartiles (covered later).

Page 5
 Number of Relative % Relative Cumulative
Frequency
Matches Frequency Frequency Frequency
2
46 2 = 0.04 0.04 x 100=4% 2
50
47 3 0.06 6% 2+3=5
48 5 0.1 10% 5+5=10
49 6 0.12 12% 10+6=16
50 16 0.32 32% 16+16=32
51 6 0.12 12% 32+6=38
52 4 0.08 8% 38+4=42
53 3 0.06 6% 42+3=45
54 3 0.06 6% 48
55 2 0.04 4% 50
Total 50 1.00 100%

The accuracy of the calculations can be checked by:

1. The total of the Relative Frequencies should be 1
2. The total of the %Relative Frequencies should be 100%
3. The last Cumulative Frequency should be equal to the total of the frequency column.

From this table, answer to questions expressed in certain ways can be obtained.

For example:

1. How many boxes contained 51 matches? The frequency for this was 6, so there were 6 boxes that
contained 51 matches.

2. For what percentage of boxes contained 50 matches? 32% of the values were 50.

3. How many boxes of matches contained 50 or less?

The cumulative frequency for 50 is the total of the frequencies including the frequency for 50. This is
32. So 32 boxes contained 50 or less matches.

4. What proportion of boxes contained 49, 50 or 51 matches?

This is found by adding up the relative frequencies for those numbers of matches. So the proportion is
0.12 + 0.32 + 0.12 = 0.56, this can also be expressed as a percentage, 56%.

5. How many matches were in the 27th box?

The frequency distribution table has ordered the data. For the number of boxes containing 50 matches,
the cumulative frequency is 32. The previous cumulative frequency is 16. This means that the 17th
(the one after the 16th) box through to the 32nd box contains 50 matches. As the 27th box is between
the 16th and the 32nd box, the 27th box will contain 50 matches.

Page 6
Grouped Frequency Distribution Tables
When the difference between the lowest and highest scores is larger, groups may have to be used. Consider
the following table for data for amount spent by seventy children at a recent show.

This data are continuous.

Forming groups for continuous data is similar to those below, that is, in the form of:

Lower amount but less than higher amount

or

lower amount ≤ x < higher amount

When deciding on groups it is important that:

(i) Each value will be placed in one group only.

(ii) There should be enough groups to include all values.
(iii) There should be between 5 and 15 groups. Use practical divisions based on the data.
(iv) Each group should contain the same number of values .
In the table below; the upper boundaries subtract lower boundaries give exactly the same
value, for example, in the first group $10 - $0 = $10, the second group $20 - $10 = $10 and
so on.
Although the upper boundary is always worded 'but less than n.n', in a practical sense it is
suggested that the upper boundary is so close to n.n that it is taken as n.n.

A Grouped Frequency Distribution Table has an extra column added.

The Group Midpoint is the numerical middle value, found by:

( lower amount + upper amount )

GroupMidpoint =
2

This value will be used in future topics.

The Grouped Frequency Distribution Table below was derived from a list of 70 values less than $100. The
groups were formed and the frequency of each group recorded in the frequency column. The next four
columns were calculated from the groups or their frequencies.

Page 7
 Amount Group Relative % Relative Cumulative
Frequency
Spent$ Midpoint Frequency Frequency Frequency
$0 but less 2
2 5 = 0.029 0.04 x 100=4% 2
than $10 70
$10 but less
3 15 0.043 4.3% 2+3=5
than $20
$20 but less
5 25 0.071 7.1% 5+5=10
than $30
$30 but less
4 35 0.057 5.7% 14
than $40
$40 but less
2 45 0.029 2.9% 16
than $50
$50 but less
8 55 0.114 11.4% 24
than $60
$60 but less
14 65 0.2 20% 38
than $70
$70 but less
18 75 0.257 25.7% 56
than $80
$80 but less
12 85 0.171 17.1% 68
than $90
$90 but less
2 95 0.029 2.9% 70
than $100
Total 70 1.00 100

When a table like this is given, there are no original data! The original values are lost into the groups. This
is the big disadvantage of Grouped Frequency Distribution Tables. The only assumption that can be made
about the original values is that they are evenly spread throughout the group.

For example; the 18 values in the group ‘$70 but less than $80’ are assumed to be equally spaced out
between $70 and $80. The higher the number of children surveyed, the more likely this is to be true. This
idea is used when calculating some statistical measures in future topics.

If the data are discrete, the construction of the table is a little different.

The addition of a column labelled Group Boundaries is required for the construction of a frequency
histogram and polygon (a graph).

Page 8
The data below are the numbers of mp3 players sold per day during a 60 day sale.

Number Group Class Relative % Relative Cumulative

Frequency
Sold Boundaries Midpoint Frequency Frequency Frequency
1 - 10 0.5 - 10.5 6 5.5 0.1 10% 6
11 - 20 10.5 - 20.5 10 15.5 0.167 16.7% 16
21 - 30 20.5 - 30.5 15 25.5 0.25 25% 31
31 - 40 30.5 - 40.5 12 35.5 0.2 20% 43
41 - 50 40.5 - 50.5 9 45.5 0.15 15% 52
51 - 60 50.5 - 60.5 6 55.5 0.1 10% 58
61 - 70 60.5 - 70.5 2 65.5 0.033 3.33% 60
Total 60 1.00 100
From this table, answer to questions expressed in certain ways can be obtained.
1. On how many days were 15 mp3 players sold?
Because the original data were lost it is not possible to determine this.

2. On how many days were 11 - 20 mp3 players sold?

This occurred on 10 days.

3. On how many days were 40 or less mp3 players sold?

The cumulative frequency for 31 - 40 is the total of the frequencies including the frequency for 31 -
40. This is 43. So on 43 days 40 or less mp3 players were sold.

4. On what proportion of days were 21 – 60 players sold?

This is found by adding up the relative frequencies for those numbers of players. So the proportion is
0.25 + 0.2 + 0.15+ 0.1 = 0.7, this can also be expressed as a percentage, 70%.

5. How many players sold per day are represented by the 27th score (when put in number sold order)?
The frequency distribution table has ordered the data. For the group 21 - 30, the cumulative frequency
is 31. The previous group’s cumulative frequency is 16. This means that the 17th (the one after the
16th) box through to the 31st box between 21 – 30 days. As the 27th score is between the 16th and the
31st score, the 27th score will be between 21 – 30 players sold. A more detailed look at this will occur
later in this module.

Stem and Leaf Plots

A stem and leaf plot is a way of ordering data by the magnitudes of the values, creating a graphical representation
of the spread of data. A table is constructed with one column for the 'stem' which is the first digit (or digits) in a
number - for example '1' from the number '10', and a second column for the 'leaf' which is the second digit/s of
the number - in this case the '0' from '10'. Raw data are entered by finding the appropriate 'stem' row, and
recording the 'leaf' digit in the leaf column for that row.
The ages of the 40 people can be displayed in a stem and leaf plot. The raw data for the ages are:

18 22 41 19 30 31 27 20
32 27 31 25 35 24 19 40
35 32 44 37 17 20 45 32
23 27 34 19 47 33 24 41
39 26 29 30 44 24 28 32

Page 9
If the numbers are entered from the table above starting with the first row, working from left to right, the
Stem and Leaf will be:

Stem Leaf
1 8 9 9 7 9
2 2 7 0 7 5 4 0 3 7 4 6 9 4 8
3 0 1 2 1 5 5 2 7 2 4 3 9 0 2
4 1 0 4 5 7 1 4
3|2 means 32 years old

When constructing a Stem and Leaf Plot, make sure that the numbers are equally spaced. The length of each
leaf can give some general information about the data. Each table should have a statement that gives place
value to the data. Once obtaining this Stem and Leaf plot, it is now useful to order each of the leaves
(leafs!).

The Ordered Stem and Leaf Plot is:

Stem Leaf
1 7 8 9 9 9
2 0 0 2 3 4 4 4 5 6 7 7 7 8 9
3 0 0 1 1 2 2 2 2 3 4 5 5 7 9
4 0 1 1 4 4 5 7
3|2 means 32 years old

Note: Stem And Leaf Plots are presented with leaves ordered – this is convention.

The advantages of a Stem and Leaf plot over a frequency distribution table are:

(i) The data are grouped and the original values are retained.

(ii) The data are in numerical order from the lowest value (top row on left) to the highest value
(bottom row on right). This will be very useful later on when calculating 5 number summaries.

If there is concern about the number of groups, that is, not enough groups, then each group can be
subdivided into two groups. There are various ways this is done, but the system used here is to replace the
existing 2 stem (values from 20 - 29) with the stems: 2 (values 20 - 24) and 2* (values 25 - 29).

Page 10
Now the data are spread over 7 groups instead of 4.

Stem Leaf
1* 7 8 9 9 9
2 0 0 2 3 4 4 4
2* 5 6 7 7 7 8 9
3 0 0 1 1 2 2 2 2 3 4
3* 5 5 7 9
4 0 1 1 4 4
4* 5 7
3|2 means 32 years old

Video ‘Organising Data using Tables’

Page 11
 Activity
1. The table of data below represents the speed of 40 cars passing a school at 9am on a school day.
Car Speed Data (in km/hr)
12 41 44 45 28 40 32 62 46 25
31 35 31 20 59 27 49 19 58 38
22 50 46 14 33 48 25 32 52 69
40 52 57 27 61 42 39 64 52 27

(a) Enter the data into a grouped frequency distribution table. Include a cumulative frequency
column. Make the first ‘group 10 to but less than 20’.
(b) Enter the data into a Stem and Leaf Plot.
(c) What percentage of cars were doing 40km/hr or more?

2. The number of students attending a class (maximum 25) for 30 lessons is given in the table below:
Students Attending Class
25 24 24 25 24 23
25 24 23 24 25 25
24 25 20 23 25 24
22 24 25 24 23 21
25 23 24 25 24 22

(a) Are the data discrete or continuous?

(b) Enter the data into a frequency distribution table (groups not required). Include a relative
frequency and % relative frequency column.
(c) What proportion of lessons contained 22 students?
(d) What percentage of lessons were fully attended?

3. The systolic blood pressures in mmHg (this is the higher value of the two blood pressure figures) of 30
patients are given in the table below.
Systolic Blood Pressures of 35 patients at a Cardiac Clinic
122 175 114 92 128 155
138 115 88 134 141 146
112 124 107 121 118 145
126 188 134 110 122 139
133 149 120 102 95 109
144 127 143 161 137

(a) Enter the data into a Stem and Leaf Plot. Use the key: 11|4 means 114.
(b) If hypertension (high blood pressure) is defined by a systolic blood pressure 140 or above, what
percentage of this group are suffering hypertension?

Page 12
4. The Shot Put distances thrown by 27 world champion shot putters are given in the table below. The
unit is metres (m).

22.25 20.19 21.39 21.25 21.19 22.07

21.72 20.37 20.45 23.09 21.19 21.22
21.07 21.55 23.12 20.91 22.58 21.97
20.22 20.38 22.37 22.19 21.70 20.54
22.67 21.58 21.72

(a) Enter the data into a Frequency Distribution Table. Your FDT must have at least 5 groups.
(b) How many have thrown less than 22m? (Use cumulative frequency to answer this)
(c) What percentage threw 21 to but less than 22m? (Use a % Relative Frequency Column).

Page 13
 Numeracy

Please Note: The graphs in this section illustrate concepts only, not the
correct formatting for assessments. For formatting requirements
refer to the style guide for your discipline (e.g. APA7, Harvard) and
your Unit Learning Materials and/or consult your Unit Assessor

Topic 2: Graphs
The graphs you use to present information depend upon the nature and the type of data.

Bar / Column Graphs

Categorical and discrete data can be displayed effectively in bar or column graphs.

Below are two graphs; a bar graph for the gender breakdown in a Lecture Group and a column graph for the
number of matches in 50 boxes (covered previously). These were drawn using Excel. Notice the equal
spacing and thickness of the bars. The spacing of bars reflects the nature of discrete data.

Gender Mix in a Frequency of Number

Lecture Group of Matches in 50
boxes
Female 20

15
Gender

Frequency

10
Male
5

0
0 5 10 15 20 25 30 46 47 48 49 50 51 52 53 54 55
Frequency Number of matches

Histograms
Numerical data can be used to construct a histogram. A histogram is basically a column graph with the bars
joined together. This reflects the nature of continuous data, however, discrete data can also be graphed as a
histogram

The graph below was drawn using Excel. Frequency is shown on the vertical axis. The horizontal axis
shows the amount spent at the show and is labelled with the class mark of each group. This is not ideal. The
preferred way to label the axis is to show the boundaries for each group on the bar, this is also shown below.

This type of graph is called a 'Frequency Histogram'.

Page 14
[last edited July 2024]
Student Learning Zone +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone
 Spending Money at a Show
20
18
16
14
12
Frequency

10
8
6
4
2
Preferred labelling
0
Excel labelling 0 10 20 30 40 50 60 70 80 90 100
5 15 25 35 45 55 65 75 85 95
$Amount

A line called the frequency polygon can be drawn on the frequency histogram.

From this it is possible to comment on the shape of the distribution.

The line is drawn from the centre of each column to the centre of the next. It should start from the
horizontal axis from the centre of an imaginary group below and extend back to the horizontal axis to the
centre of an imaginary group above.

The polygon can be drawn with or without the histogram present. The polygon is drawn below.

Spending Money at a Show

20
18
16
14
12
Frequency

10
8
6
4
2
0
0 10 20 30 40 50 60 70 80 90 100 110

$Amount

Another type of graph is based on the Cumulative Frequency Histogram.

A line is draw on this called an Ogive.

The Ogive can be drawn with or without the histogram.

Page 15
The Ogive is drawn from the previous cumulative frequency on the left of the column to the current
cumulative frequency on the right of the column.

Spending Money at a Show

80

70

60

50
Cumulative Frequency

40

30

20

10

0
0 10 20 30 40 50 60 70 80 90 100
5 15 25 35 45 55 65 75 85 95
$Amount

Page 16
Line Graphs
Another important type of graph is the Line Graph.

A line graph shows change of a variable (usually) over a period of time.

Because none of the data collected for our scenario is suitable for a line graph, a simple example has been
made up.

Year % Pass Percentage Pass Rate for Unit Code

Rate ABC12345
100
2004 72 80

% Pass Rate
2005 79 60
2006 77 40
2007 82 20

2008 85 0
2004 2005 2006 2007 2008 2009
2009 87 Year

Pie or Sector Graphs

The next graph is very useful when you are trying to show how a whole set of data is divided up into
components. For example: If you collected data about how students travel to university, the whole group can
be divided up into those that travel by bus, car, bike, walk etc. A pie graph is good at showing the whole in
its individual parts.

Method of Travel to University by Students.

bike
other
8%
12%

Car
50%
bus
18%
walk
12%

Visually, it is easy to say that most students travel to uni by car. With the percentages given, it is also
possible to quantify information from the graph.

For example: If there are 2000 students on campus today, approximately how many travelled by bike?
Number travelling by bike =12% of 2000 = 0.12 x 2000 = 240 students.

A variation of the pie graph is the percentage bar graph. Ideally a bar of length 100mm (or 200mm, 300mm
etc.) is divided by in the percentages given.

Page 17
 Car 50% Bus 18% Walk Bike Other
12% 8% 12%

Composite Bar / Column Graph

Another graph to consider is a composite bar graph.

This graph can be drawn either vertically or horizontally.

It can be used to compare two or more sets of similar data.

The graph below compares average weekly earnings in May 2007 for males and females working full time
in different states and territories.

Average Weekly Earnings May 2007

1600
1400
Average Weekly Earnings $

1200
1000
800
Male
600
Female
400
200
0
NSW Vic Qld SA WA Tas NT Act
Location

Data sourced from:

[Link]
573D20010F2DD?opendocument

Video ‘Presenting Data in Graphs’

Page 18
Activity
1. The height of a plant is measured every Friday morning for twelve weeks. Which type of graph would
be best to show the growth of the plant?
2. A ‘sound and vision’ shop sells CDs, DVDs, console games and computer games. Which type of
graph would best display the relative sales of the different items sold?
3. A student wishes to compare the number of motorcycle fatalities between the states. To get a more
accurate picture, data are collected for the years 2007, 2008 and 2009. Which type of graph allows
all data to be presented in one graph?
4. The number of students attending a class (maximum 25) for 30 lessons is given in the table below:
Number attending Tally Frequency
20 Ι 1
21 Ι 1
22 ΙΙ 2
23 ΙΙΙΙ 5

24 ΙΙΙΙ ΙΙΙΙ Ι 11

25 ΙΙΙΙ ΙΙΙΙ 10
Total 30
Construct a column graph to represent this data.

5. The Shot Put distances thrown by 27 world champion shot putters are given in the Frequency
Distribution Table below. The unit is metres (m).
Distance Tally Frequency Cumulative % Relative
Frequency Frequency
5
20 to but less than 20.5 ΙΙΙΙ 5 5 × 100 =
18.5%
27
2
20.5 to but less than 21 ΙΙ 2 7
27
× 100 =
7.4%

6
21 to but less than 21.5 ΙΙΙΙ Ι 6 13 × 100 =
22.2%
27
6
21.5 to but less than 22 ΙΙΙΙ Ι 6 19
27
× 100 =
22.2%

4
22 to but less than 22.5 ΙΙΙΙ 4 23
27
× 100 =
14.8%

2
22.5 to but less than 23 ΙΙ 2 25
27
× 100 =
7.4%

2
23 to but less than 23.5 ΙΙ 2 27
27
× 100 =
7.4%

Total 27 100%

(a) Construct a frequency histogram for this information. As a second step, put a frequency polygon
on the histogram.
(b) Construct a cumulative frequency histogram and then add an Ogive to the histogram.

Page 19
Page 20
 Numeracy

Topic 3: Measures of Central Tendency

Measures of Central Tendency are summary statistics that attempt to represent the data by summarising them
as a single, 'typical' value. There are three commonly used measures to describe the 'central' or 'typical' value,
the Mean, Median and Mode.

Mean
The mean is commonly referred to as the average.

The mean is the most common measure used. It is found by adding up all the values and dividing by the
number of values. Because of this, every value contributes to the mean.

For a set of values written as X1, X2, X3, X4, X5,………. Xn, the sample mean is calculated using the equation:

=X =
sum of the values ∑X
number of values in the sample n

where X is the mean of the n values in the sample.

The symbol Σ is shorthand for "the sum of".

The equation for the population mean uses some different pronumerals.

=µ =
sum of the values ∑X
number of values in the population N

where μ is the mean of the N values in the population.

Because the calculation of the sample and population means is the same (except for the pronumerals used in
the equation), calculators have just one key to cover both.

Page 21
Student Learning Zone +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone [last edited July 2024]
Finding the mean of Individual Values
These data are the maths test results of a sample of 15 students in ascending order.

23, 45, 50, 54, 55, 57, 59, 59, 59, 61, 63, 75, 75, 81, 90.

X=
∑X
n
23 + 45 + 50 + ............. + 75 + 81 + 90
X=
15
X = 60.4
The mean can easily be calculated on a calculator, especially a scientific calculator with STATS mode
(covered later).

Finding the mean from a Frequency Distribution Table

Let’s recall the data about the number of matches in 50 boxes which was organised in a FDT with no
grouping. The table has a new column added headed f x x. This column is the score multiplied by the
frequency.

Number of f ×X 92 represents the total of all the

Frequency (f) 46s in the data.
Matches (X)
141 represents the total of all
46 2 2 x 46 = 92 the 47s in the data.
3 x 47 =
47 3
141
48 5 240
49 6 294
50 16 800
51 6 306
52 4 208
53 3 159
54 3 162
2512 is the total of all the
values in the data.
55 2 110
Total Σf =50 ΣfX =
2512

The mean number of matches per box is given by the equation:

sum of the values
X=
number of values in the sample
ΣfX
X=
Σf
2512
X=
50
X = 50.24
The mean number per box is 50.24 matches.

Now let’s look at the example about the money spent by children at the show. This was organised using a
Grouped FDT.
Page 22
Finding the mean is performed in a similar manner to above except the Group Midpoint is used to represent
the group.

Group f ×X
Frequency
Amount Spent$ Midpoint
(f)
(X)
$0 but less than $10 2 5 2 x 5 = 10
$10 but less than $20 3 15 3 x 15 = 45
$20 but less than $30 5 25 125
$30 but less than $40 4 35 140
$40 but less than $50 2 45 90
$50 but less than $60 8 55 440
$60 but less than $70 14 65 910
$70 but less than $80 18 75 1350
$80 but less than $90 12 85 1020
$90 but less than 190
2 95
$100
Total Σf =70 ΣfX =
4320

The mean amount spent is given by the equation:

sum of the values

X=
number of values in the sample
ΣfX
X=
Σf
4320
X=
70
X = $61.71 (to the nearest cent)

Remember this method assumes that the values in each group are evenly spread. This assumption is not
always true so the figure obtained for the mean using this method can slightly different to the mean if the
original values where used.

Finding the mean using a Calculator

A scientific calculator is usually capable of performing this operation when STATS mode is used.

The instructions below apply to the Casio fx-82AU and the Sharp EL531 only. If you have a different
scientific calculator to this, you will need to read the user's guide for your calculator.

There are 3 steps to using a scientific calculator in STATS mode:

1. Getting the calculator into STATS mode.

2. Entering the values (data).

Page 23
 3. Obtaining the statistical measure required.

On a Casio fx-82AU: On a Sharp EL531

To get this calculator into STATS mode To get this calculator into STATS mode
Press Mode . The display will be: Press Mode .
Press 1 for STAT
Press 0 for Standard Deviation (+ more)

The calculator should display Stat 0 in the top left of

the display.

Press 2 for STAT

Press 1 for 1-VAR

The calculator should display a small STAT on the top

line of the display.

Now your calculator is in STATs mode, the next stage is to enter the data. There are slight differences
depending on how the data are organised.

Finding the Mean of Individual Values

These data are the maths test results of a sample of 15 students in ascending order.

23, 45, 50, 54, 55, 57, 59, 59, 59, 61, 63, 75, 75, 81, 90.

On a Casio fx-82AU: On a Sharp EL531

Each number is entered into the list, 23 =, 45 =: Each number from the list is entered as:
23 then press M+
DATA SET =
the calculator display will show 1
45 then press M+
DATA SET=
the calculator display will show 2
.
.
.
.
.

Page 24
 If you make a mistake, use the big round blue ‘Replay’
button to navigate back to the position of the incorrect
90 then press M+
value and enter the correct value. the calculator display will show
DATA SET=
15
DATA SET=
The display 15 informs you that 15
numbers have been entered into the STATs memory.

If you make a mistake and need to start again, the

STATs memory is cleared by pressing
2ndF CA

Now the data are entered, the calculated value of the sample and population standard deviations can be
obtained.

On a Casio fx-82AU: On a Sharp EL531

When all the values have been entered, press AC To obtain the value of the population standard
deviation; press
This is important.
RCL σ x the symbol σ is used to represent the
Now press SHIFT 1 . This is the same as population standard deviation.
SHIFT STAT .
To obtain the value of the sample standard deviation;
press
The calculator display will be as shown below.
RCL sx the symbol s is used to represent the
sample standard deviation.

(Think of RCL as recall)

Other values are found by;

To obtain the value of the mean; press

RCL x the symbol x means the mean.

To obtain the number of values entered; press

RCL n the symbol n means the number of numbers
entered.

To obtain the sum of the values entered; press

RCL Σx the symbol Σx means the sum of the
numbers entered.
Now press 4 to get more options.
To obtain the sum of the squares of the values entered;
press
RCL Σx 2 the symbol Σx 2 means the sum of the
squares of the numbers entered.

Press 2 = – for the mean

Press 3 = – for the population standard dev.
Press 4 = – for the sample standard dev.
Always press = to get the value required.

Do the calculation yourself, the values obtained should be:

Mean = 60.4

Page 25
Finding the Mean from a Frequency Distribution Table
Let’s recall the data about the number of matches in 50 boxes which was organised in a FDT with no
grouping.
Number of
Frequency (f)
Matches (X)
46 2
47 3
48 5
49 6
50 16
51 6
52 4
53 3
54 3
55 2
Total Σf = 50

It is worth recalling that this table is informing us that there are 2 occurrences of the value 46, 3 occurrences
of 47, 5 occurrences of 48, ….
Instead of entering the value of 50, 16 times a modified entry process is used. The process for finding the
standard deviation is exactly the same.

Remember to clear STATs memory first.

With a Frequency Distribution Table, the data are entered into the Sharp EL531 calculator in the format:
score, frequency .

On the Casio fx-82AU, the table the values are entered into needs to be changed into a frequency like the
example above. To do this:

Press SHIFT MODE . Eight options will be displayed.

Press the big blue key down .

This will take you to another screen with 5
options.

Press 3: STAT The screen will then display

Press 1: ON This will add a frequency column

Page 26
 On a Casio fx-82AU: On a Sharp EL531
Get the calculator into STATS mode: Each row from the table is entered as:
Press Mode . The display will be: 46 , 2 M +
DATA SET =
the calculator display will show 1
47 , 3 M +
DATA SET=
the calculator display will show 2

DATA SET=
The display 2 informs you that 2
rows have been entered into the STATs memory.
Press 2 for STAT

48 , 5 M +
DATA SET=
the calculator display will show 3
.
.
.
.
.
Press 1 for 1-VAR .
.
This time the calculator will display an X column and
a FREQ column. Enter a value followed by =.
55 , 2 M + the calculator display will show
Navigate to the frequency column using the big blue DATA SET=
REPLAY key. 15

If you make a mistake and need to start again, the

STATs memory is cleared by pressing
2ndF CA

Obtaining values for the mean is exactly the same process as in the previous section.

Do the calculation yourself, the values obtained should be:

Mean = 50.24

Page 27
If the data are grouped, the process is almost identical except that the group midpoint is used.
Group
Frequency
Amount Spent$ Midpoint
(f)
(X)
$0 but less than $10 5 2
$10 but less than $20 15 3
$20 but less than $30 25 5
$30 but less than $40 35 4
$40 but less than $50 45 2
$50 but less than $60 55 8
$60 but less than $70 65 14
$70 but less than $80 75 18
$80 but less than $90 85 12
$90 but less than
95 2
$100
Total Σf =70

The data are entered into the calculator as Group Midpoint, Frequency or Group Midpoint; Frequency
depending on the calculator used.

Median
The median is the middle value of the data if the values are arranged in order.

As a measure of centre, it is a value based on its position; it is not influenced by the size of the values that
are above or below it.

Therefore it is quite different to the mean because it is based on position and for the mean, every value
contributes.

Finding the median of Individual Values

The following data are the maths test results of a sample of 15 students in ascending order.

23, 45, 50, 54, 55, 57, 59, 59, 59, 61, 63, 75, 75, 81, 90.

When finding the median for a set of individual values, the first step is to order the values. Usually this is
done in ascending order, but descending order gives the same result.

The position of the median can always be found by using a simple expression:
n +1
The position of the median is .
2

As there are 15 values, the position of the median is:

n + 1 15 + 1 16
= = = 8
2 2 2
th
The 8 value is the median.

Page 28
The median is the 8th value, which is 59. The median is the 8th value, so there are 7 values before it and 7
values after it.

23, 45, 50, 54, 55, 57, 59 59 59, 61, 63, 75, 75, 81, 90
7 values median 7 values = 15 values

The median cannot be calculated on a scientific calculator.

Finding the median from a Frequency Distribution Table

Let’s recall the data about the number of matches in 50 boxes which was organised in a FDT with no
grouping.

To find the median, the cumulative frequency column is required.

Number of Cumulative
Frequency (f)
Matches (X) Frequency
46 2 2
47 3 5
The 17th through to the 32nd
48 5 10 values are in this group.

49 6 16 This means that the 25th and

26th values are both 50.
50 16 32
51 6 38 The median is 50.

52 4 42
53 3 45
54 3 48
55 2 50
Total Σf =50

n + 1 50 + 1 51
As there are 50 values, the position of the median is: = = = 25.5
2 2 2

In this case the 25th and 26th values are required. The median is the average of the 25th and 26th values. As
the 25th and 26th values are both 50, the median is 50.

Now let’s look at the example using a Grouped FDT. Finding the median can be performed in one of two
ways: interpolation method or graphical method.

Page 29
(a) Interpolation Method (Amount Spent at a Show Data)

Group Cumulative
Frequency
Amount Spent$ Midpoint Frequency
(f)
(X)
$0 but less than $10 2 5 2
$10 but less than $20 3 15 5
$20 but less than $30 5 25 10
$30 but less than $40 4 35 14
This means that the 24th
$40 but less than $50 2 45 16 score is $60
$50 but less than $60 8 55 24
$60 but less than $70 14 65 38 This means that the 38th
$70 but less than $80 18 score is $70
75
There are 14 values in
56
The group width
$80 but =70-60
less than $90 12 this group.
85 68
=10
$90 but less than
2 95 70
$100
Total Σf =70
n + 1 70 + 1 71
As there are 70 values, the position of the median is: = = = 35.5
2 2 2

In this case the 35th and 36th values are required. Both of these values are located in the group ‘$60 but less
than $70’. This means that the median is greater than 60 but less than 70. This means that the median is
(35.5 – 24 =) 11.5/14 of the way through the ‘$60 but less than $70’ group. The median is:

Median = 60 +
( 35.5 − 24 ) ×10 (Group width)
14
Median = 68.2

Or expressed more generally:

 n +1
  -cf m −1
Lm + 
2 
Median = × Group width
fm
where Lm is the lower limit of the median group
f m is the frequency of the median group
cf m −1 is the cumulative frequency of the previous group

(b) Graphical method (using the Ogive)

There are 70 values in the table. This is shown on the vertical axis (Cumulative Frequency).

The first step is to come half way up the vertical axis.

This is to 35.

Page 30
The second step; from this position, move horizontally across the graph until the cumulative frequency
Ogive is found.

The third step is to move down (vertically) to the horizontal axis and read the value.

Ogive of Money Spent

80

70

60
Cumulative Frequency

50

40

30

20

10

0
0 20 40 60 80 100 120
Amount Spent $

From the graph, the median is approximately 68.

Remember this method assumes that the values in each group are evenly spread. This assumption is not
always true, so the figure obtained for the median using this method can slightly different to the actual
median if the original values where used.

Page 31
Mode
The Mode is the value that occurs the most often.

The mode may not exist because values only occur once.

The mode could even be quite different to the mean and median, it may be much higher or much lower
depending on the meaning of the data.

There may also be more than one mode. If there are two modes, the data are said to be ‘bimodal’.

Finding the mode from Graphs

Number of Children in 40 households

16
14
12
10
Frequency

8
6
4
2
0
0 1 2 3 4 5
Number of Children

From the graph, it can be observed that '1 child per household' has the highest frequency. This is the mode of
these values.

Finding the mode of Individual Values

The following data are the maths test results of a sample of 15 students in ascending order.

23, 45, 50, 54, 55, 57, 59, 59, 59, 61, 63, 75, 75, 81, 90.

When inspecting this data it can be seen that the 59 occurs 3 times.

This is the value that occurs the most frequently, that is, it has the highest frequency.

The mode for this data is 59.

Finding the mode from a Frequency Distribution Table

Let’s recall the data about the number of matches in 50 boxes which was organised in a FDT with no
grouping.

To find the mode, the frequency column is inspected. The highest frequency of 16 is related to a box
containing 50 matches. The mode of this data is 50 matches. Take care to give the mode as the score (50
matches) not the frequency (16).

Page 32
 Number of
Frequency (f)
Matches (X)
46 2
The mode for this
47 3 data is 50 matches as
it has the highest
48 5 frequency (occurs
the most often)
49 6
50 16
51 6
52 4
53 3
54 3
55 2
Total Σf =50

When data are grouped, it is more relevant to refer to the modal group.
From the scenario, ‘Amount Spent at the Show’, the group with the highest frequency is '$70 but less than
$80' which has a frequency of 18.

It can be stated that the modal group is '$70 but less than $80'.

Group
Frequency
Amount Spent$ Midpoint
(f)
(X)
$0 but less than $10 2 5
$10 but less than $20 3 15
$20 but less than $30 5 25
$30 but less than $40 4 35
The modal group for
$40 but less than $50 2 45 this data is ‘$70 but
$50 but less than $60 8 55 less than $80’ as it
has the highest
$60 but less than $70 14 65 frequency (occurs
the most often)
$70 but less than $80 18 75
$80 but less than $90 12 85
$90 but less than
2 95
$100
Total Σf =70

Either the group ‘$70 but less than $80’ or the Group Midpoint ‘$75’ can be given as the mode.

Page 33
Which measure of centre should be used and when?
Every value from the data is used to calculate the mean. As long as the data are spread evenly, without excessive
variation, the mean is usually used. A lecturer would use the mean to find a measure of centre because the
marks would be usually spread out from (say) 30 to 100%. In this data the low or high values are not
excessively different to a measure of centre.

In real estate, the median is often used as a measure of centre because there are some excessively high
values that are quite different to the bulk of the market. Real estate sales generally consist of many
properties at the cheaper part of the market and fewer properties at the expensive part of the market. If the
mean was to be used, the fewer, more expensive properties would have a huge effect on the mean. Because
the median is a measure of centre based on location, variation in the price of high value properties and the
number of high value properties will have no effect on the median but would have a significant effect on the
mean.

In the clothing industry, the mode is often used as a measure of centre. If a shop sells mostly size 12
clothing, then size 12 is the mode of the sizes sold. The mean is of little use because it could be a value that
is not even a possible size. The median is a better measure but fails to reflect the nature of the sales
environment.

Video ‘Measures of Central Tendency’

Page 34
Activity

1. The lifetime, in hours, of a sample of 15 light bulbs is: 351, 429, 885, 509, 317, 753, 827, 737, 487,
726, 395, 773, 926, 688, 485.

Calculate the mean, mode and median of the values. Do not organise into a table.

2. The number of children in a 10 families is: 1, 5, 2, 2, 2, 3, 1, 4, 3, 2. Calculate the mean, mode and
median of the values. Do not organise into a table.

3. For the Car Speed Data, calculate the mean, mode and median of the values after organising the data
in Frequency Distribution Table (The FDT can be found in the answers to the Topic ‘Graphs’).

Car Speed Data (in km/hr)

12 41 44 45 28 40 32 62 46 25
31 35 31 20 59 27 49 19 58 38
22 50 46 14 33 48 25 32 52 69
40 52 57 27 61 42 39 64 52 27

4. The number of students attending a class (maximum 25) for 30 lessons is given in the table below:

Students Attending Class

25 24 24 25 24 23
25 24 23 24 25 25
24 25 20 23 25 24
22 24 25 24 23 21
25 23 24 25 24 22

(a) Calculate the mean, mode and median using a Frequency Distribution Table. (The FDT can be
found in the answers to the Topic ‘Graphs’).

(b) Discuss the appropriateness of each measure of central tendency as a typical value.

5. The Shot Put distances thrown by 27 world champion shot putters are given in the table below. The
unit is metres (m).

22.25 20.19 21.39 21.25 21.19 22.07

21.72 20.37 20.45 23.09 21.19 21.22
21.07 21.55 23.12 20.91 22.58 21.97
20.22 20.38 22.37 22.19 21.70 20.54
22.67 21.58 21.72

Calculate the mean, mode and median using a grouped Frequency Distribution Table. (The FDT can
be found in the answers to the Topic ‘Graphs’)

Page 35
 Numeracy

Topic 4: Measures of Spread

Consider the two sets of values below. They both have a mean and median of 50, but they are quite
different.

Set A: 20, 35, 50, 65, 80 Set B: 40, 45, 50, 55, 60

The reason the sets are different is because of the spread of the values; the values in Set A are more spread
out than those in Set B.

Range
In statistics, not only is a measure of centre important but a measure of spread is also important. The most
basic measure of spread is the range.

Range = Highest Value - Lowest Value

For the data above,

Set A Set B
Range = Highest Value - Lowest Value Range = Highest Value - Lowest Value
Range = 80 – 20 Range = 60 – 40
Range = 60 Range = 20

From this it is possible to say that Set A is has a higher spread than Set B. The problem with the range is that
it uses the lowest and highest values. In any set of data, the lowest or highest score could be an odd or
unusual value.

Using odd or unusual values to measure spread will not produce a result that properly reflects the data. If
you consider students sitting an examination, a student not feeling well sits an exam and records an
unusually low score. This score is not typical of the group and so the range is much larger than it should be
for the group.

Scores that are atypical of the group are called outliers. There are ways of identifying outliers, but these will
not be covered in this unit.

Page 36
Student Learning Zone +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone [last edited June 2023]
Finding the range of Individual Values
The following data are the maths test results of a sample of 15 students in ascending order.

23, 45, 50, 54, 55, 57, 59, 59, 59, 61, 63, 75, 75, 81, 90.

Once the data are ordered, the lowest and highest values are easy to locate.

Range = Highest Value - Lowest Value

Range = 90 – 23
Range = 67

Finding the range from a Frequency Distribution Table

Let’s recall the data about the number of matches in 50 boxes which was organised in a FDT with no
grouping. To find the range, the ‘Number of Matches’ column is used.

Number of Frequency (f) The lowest score obtained

Matches (X) was 46. The highest
score obtained was 55.
46 2 Range = 55 – 46 = 9

47 3
48 5
49 6
50 16
51 6
52 4
53 3
54 3
55 2
Total Σf =50

When data are grouped it is more difficult to determine the highest and lowest values so an assumption must
be made.
Group
Frequency
Amount Spent$ Midpoint
(f)
(X)
$0 but less than $10 2 5
$10 but less than $20 3 15
$20 but less than $30 5 25
$30 but less than $40 4 35
$40 but less than $50 2 45
$50 but less than $60 8 55
$60 but less than $70 14 65
$70 but less than $80 18 75
$80 but less than $90 12 85

Page 37
 $90 but less than
2 95
$100
Total Σf =70

The only assumption (because the original values are not present) that can be made here is that the lowest
value is $0 and the highest value is $100.
The range is 100 – 0 = 100

Interquartile Range
The next measure of spread is the Interquartile Range (IQR). It is still a range but it is between the first and
third quartiles. The first and third quartiles are values that are ¼ and ¾ of the way through the ordered data.
Another way to think about the IQR is to consider it the range of the middle 50% of the data.

Diagrammatically, it can be represented as below:

Lowest First Second

Quartile Third Highest
Value Quartile Value
or Quartile
Median

Inter-Quartile Range = Third Quartile - First Quartile

Or
= Q3 − Q1
IQR

The five values shown in this diagram are commonly referred to as a 'five number summary'. A five number
summary is very useful for calculating the range, IQR and for drawing a box and whisker diagram (covered
later).

Finding the interquartile range of Individual Values

The following data are the maths test results of a sample of 15 students in ascending order.

23, 45, 50, 54, 55, 57, 59, 59, 59, 61, 63, 75, 75, 81, 90.

Once again the data must be ordered. Previously a simple equation was used to locate the median. This is
similar.

n +1 3(n + 1)
th th
The first quartile is the score. The third quartile is the score.
4 4

Page 38
For the data above:

First Quartile (Q1) Third Quartile (Q3)

n +1 3 ( n + 1)
4 4
15 + 1 3 × 16
= =
4 4
16 48
= =
4 4
= 4th value = 12th value

23, 45, 50, 54, 55, 57, 59, 59, 59, 61, 63, 75, 75, 81, 90.

4th value 12th value

= Q3 − Q1
IQR
= 75 − 54
IQR
IQR = 21

The advantage of using the IQR is that the first and third quartiles are stable values; they are not outliers or
'odd' values.

The IQR is the range of the middle 50% of the data.

Finding the interquartile range from a Frequency Distribution Table

Let’s recall the data about the number of matches in 50 boxes which was organised in a FDT with no
grouping. To find the IQR, the ‘Cumulative Frequency’ column is used.

Number of Cumulative
Frequency (f)
Matches (X) Frequency The 11th through to the 16th values are
49 matches, this includes the 12th and
46 2 2 13th values.
47 3 5
The first quartile is 49.
48 5 10
49 6 16
The 33rd through to the 38th values are
50 16 32 51 matches.
51 6 38 The 38th value is 51 matches.
52 4 42
53 3 45 The 39th through to the 42nd values are
52 matches.
54 3 48
The 39th value is 52 matches.
55 2 50
Total Σf =50

First Quartile (Q1) Third Quartile (Q3)

Page 39
 n +1 3 ( n + 1)
4 4
50 + 1 3 × 51
= =
4 4
51 153
= =
4 4
= 12.75th value = 38.25th value

To find the 12.75th value, means finding the 12th and 13th values. In this case the 12th and 13th values are both
49, so the First Quartile is 49.

To find the 38.25th value, we have to calculate the value that is 0.25th (i.e. a quarter) of the way between the
38th value and the 39th value. In this case the 38th value is 51 and the 39th value is 52, so the Third Quartile is
51+ 0.25 x Difference between the 39th and 38th values, i.e. 52-51.

This is equal to 51+ 0.25 x 1 = 51.25

= Q3 − Q1
IQR
=
IQR 51.25 − 49
IQR = 2.25

When data are grouped, the cumulative frequency is used to obtain the quartiles. There are two methods to
find the quartiles; the interpolation method and the graphical method.

(a) Interpolation Method

Group Cumulative
Frequency
Amount Spent$ Midpoint Frequency
(f)
(X)
$0 but less than $10 2 5 2
$10 but less than $20 3 15 5 The 17th through to the 24th
values are found in the group
$20 but less than $30 5 25 10 ‘$50 but less than $60’

$30 but less than $40 4 35 14 The 17.75th value is in this

group.
$40 but less than $50 2 45 16
$50 but less than $60 8 55 24
$60 but less than $70 14 65 38
$70 but less than $80 18 75 56 The 39th through to the 56th
$80 but less than $90 12 85 68 values are found in the group
‘$70 but less than $80’
$90 but less than
2 95 70 The 53.25th value is in this
$100
group.
Total Σf =70

Page 40
 First Quartile (Q1) Third Quartile (Q3)
n +1 3 ( n + 1)
4 4
70 + 1 3 × 71
= =
4 4
71 213
= =
4 4
= 17.75th value = 53.25th value

The first quartile is found in the group ‘$50 but less than $60’.

In this group there are 8 values (the 17th to the 24th) and the class width is 10.

17.75 − CFPr eviousGroup 17.75 − 16 1.75

This means the First Quartile is = = of the way through the group ‘$50 but less than
GroupFrequency 8 8
$60’.

The third quartile is found in the group ‘$70 but less than $80’.

In this group there are 18 values (the 39th to the 56th) and the class width is 10.

53.25 − CFPr eviousGroup 53.25 − 38 15.25

This means that the Third Quartile is = = of the way through the group ‘$70 but
GroupFrequency 18 8
less than $80’.

1.75 15.24
The first quartile (Q1 ) =50 + ×10 The third quartile (Q3 ) =
70 + ×10
8 18
= 52.19 = 78.47

= Q3 − Q1
IQR
=
IQR 78.47 − 52.19
IQR = 26.28

(b) Graphical Method (using the Ogive)

There are 70 values in the table. This is shown on the vertical axis.

To find the first quartile, come one quarter of the way up the vertical axis. This is to 17.5 (One quarter of
70). From this position, move horizontally across the graph until the cumulative frequency line is found,
then move down (vertically) to the horizontal axis and read the value.

To find the third quartile, come three quarters of the way up the vertical axis. This is to 52.5 (three quarters
of 70). From this position, move horizontally across the graph until the cumulative frequency line is found,
then move down (vertically) to the horizontal axis and read the value.

Page 41
 Ogive of Money Spent
80

70

Cumulative Frequency 60

50

40

30

20

10

0
0 20 40 60 80 100 120
Amount Spent $

The graph value for the first quartile is (about) 52 and the third quartile is (about) 78, giving an IQR of 78 –
52 = 26.

Standard Deviation
The third measure of spread is the Standard Deviation. The key word in this measure is deviation. The word
deviation in this context is how much each value deviates (differs) from the mean.

For example: consider the sample data set 2, 4, 6, 8.

Data Mean Deviation from Mean Square the deviations

(X − X ) (this makes the deviations positive)

(X − X )
2

2
X=
∑X 2 - 5 = -3 9
n
4 4 - 5 = -1 1
2+ 4+6+8
=
6 4 6-5=1 1
20
=
8 4 8-5=3 9
=5
∑( X − X )
2
=
20
The standard deviation of a sample is calculated using

∑ ( X −=
X)
2
20
=S = = 2.58
6.66666
n −1 4 −1

Page 42
If the set of data is a population, then the standard deviation has a slightly different equation. In this equation X
represents the raw scores, the population mean is represented by mu, while N is the population size.:

∑ ( X − µ )=
2
20
σ
= = =
5 2.24
N 4

A scientific calculator is usually capable of performing this operation when STATS mode is used.

The instructions below apply to the Casio fx-82AU and the Sharp EL531 only. If you have a different
scientific calculator to this, you will need to read the user's guide for your calculator.

There are 3 steps to using a scientific calculator in STATS mode:

1. Getting the calculator into STATS mode.

2. Entering the values (data).

3. Obtaining the statistical measure required.

On a Casio fx-82AU: On a Sharp EL531

To get this calculator into STATS mode To get this calculator into STATS mode
Press Mode . The display will be: Press Mode .
Press 1 for STAT
Press 0 for Standard Deviation (+ more)

The calculator should display Stat 0 in the top left of

the display.

Press 2 for STAT

Press 1 for 1-VAR

The calculator should display a small STAT on the top

line of the display.
Now your calculator is in STATs mode, the next stage is to enter the data. There are slight differences
depending on how the data is organised.

Finding the Standard Deviation of Individual Values

The following data are the maths test results of a sample of 15 students in ascending order.

23, 45, 50, 54, 55, 57, 59, 59, 59, 61, 63, 75, 75, 81, 90.

Page 43
 On a Casio fx-82AU: On a Sharp EL531
Each number is entered into the list, 23 =, 45 =: Each number from the list is entered as:
23 then press M+
DATA SET =
the calculator display will show 1
45 then press M+
DATA SET=
the calculator display will show 2
.
.
.
If you make a mistake, use the big round blue ‘Replay’ .
button to navigate back to the position of the incorrect .
value and enter the correct value.
90 then press M+
DATA SET=
the calculator display will show 15
DATA SET=
The display 15 informs you that 15
numbers have been entered into the STATs memory.

If you make a mistake and need to start again, the

STATs memory is cleared by pressing
2ndF CA

Now the data have been entered, the calculated value of the sample and population standard deviations can be
obtained.

On a Casio fx-82AU: On a Sharp EL531

When all the values have been entered, press AC To obtain the value of the population standard
deviation; press
This is important.
RCL σ x the symbol σ is used to represent the
Now press SHIFT 1 . This is the same as population standard deviation.
SHIFT STAT . To obtain the value of the sample standard deviation;
press
The calculator display will be as shown below.
RCL sx the symbol s is used to represent the
sample standard deviation.

(Think of RCL as recall)

Other values are found by;

To obtain the value of the mean; press

RCL x the symbol x means the mean.

To obtain the number of values entered; press

RCL n the symbol n means the number of numbers
entered.

To obtain the sum of the values entered; press

Page 44
 Now press 4 to get more options. RCL Σx the symbol Σx means the sum of the
numbers entered.

To obtain the sum of the squares of the values entered;

press
RCL Σx 2 the symbol Σx 2 means the sum of the
Press 2 = – for the mean squares of the numbers entered.
Press 3 = – for the population standard dev.
Press 4 = – for the sample standard dev.
Always press = to get the value required.

Do the calculation yourself, the values obtained should be:

Population Standard Deviation =15.4177387 rounded gives 15.4

Sample Standard Deviation = 15.95887572 rounded gives 16.0

Finding the Standard Deviation from a Frequency Distribution Table

Let’s recall the data about the number of matches in 50 boxes which was organised in a FDT with no
grouping.
Number of
Frequency (f)
Matches (X)
46 2
47 3
48 5
49 6
50 16
51 6
52 4
53 3
54 3
55 2
Total Σf = 50

It is worth recalling that this table is informing us that there are 2 occurrences of the value 46, 3 occurrences
of 47, 5 occurrences of 48, ….
Instead of entering the value of 50, 16 times a modified entry process is used. The process for finding the
standard deviation is exactly the same.

Remember to clear STATs memory first.

With a Frequency Distribution Table, the data are entered into the Sharp EL531 calculator in the format:
score, frequency .

On the Casio fx-82AU, the table the values are entered into needs to be changed into a frequency like the
example above. To do this:

Page 45
Press SHIFT MODE . Eight options will be displayed.

Press the big blue key down .

This will take you to another screen with 5
options.

Press 3: STAT The screen will then display

Press 1: ON This will add a frequency column

On a Casio fx-82AU: On a Sharp EL531

Get the calculator into STATS mode: Each row from the table is entered as:
Press Mode . The display will be: 46 , 2 M +
DATA SET =
the calculator display will show 1
47 , 3 M +
DATA SET=
the calculator display will show 2

DATA SET=
The display 2 informs you that 2
rows have been entered into the STATs memory.
Press 2 for STAT

48 , 5 M +
DATA SET=
the calculator display will show 3
.
.
.
.
.
Press 1 for 1-VAR .
.
This time the calculator will display an X column and
a FREQ column. Enter a value followed by =. 55 , 2 M + the calculator display will show
Navigate to the frequency column using the big blue DATA SET=
REPLAY key. 15

If you make a mistake and need to start again, the

STATs memory is cleared by pressing
2ndF CA

Page 46
Obtaining values for the population and sample standard deviation is exactly the same process as in the
previous section.

Do the calculation yourself, the values obtained should be:

Population Standard Deviation =2.140554106 rounded gives 2.14

Sample Standard Deviation = 2.162387192 rounded gives 2.16

If the data are grouped, the process is almost identical except that the group midpoint is used.
Group
Frequency
Amount Spent$ Midpoint
(f)
(X)
$0 but less than $10 5 2
$10 but less than $20 15 3
$20 but less than $30 25 5
$30 but less than $40 35 4
$40 but less than $50 45 2
$50 but less than $60 55 8
$60 but less than $70 65 14
$70 but less than $80 75 18
$80 but less than $90 85 12
$90 but less than
95 2
$100
Total Σf =70

The data are entered into the calculator as Group Midpoint, Frequency or Group Midpoint; Frequency
depending on the calculator used.

Video ‘Measures of Spread’

Page 47
Activity

1. The lifetime, in hours, of a sample of 15 light bulbs is: 351, 429, 885, 509, 317, 753, 827, 737, 487,
726, 395, 773, 926, 688, 485.

Calculate the range, inter-quartile range and standard deviation of the values. Do not organise into a
table.

2. The numbers of children in 10 families are: 1, 5, 2, 2, 2, 3, 1, 4, 3, 2. Calculate the range, inter-quartile

range and standard deviation of the values. Do not organise into a table.

3. For the Car Speed Data, calculate the range, inter-quartile range and standard deviation of the values
after organising the data in Frequency Distribution Table (This can be found in the answers to the
Topic ‘Graphs’).

Car Speed Data (in km/hr)

12 41 44 45 28 40 32 62 46 25
31 35 31 20 59 27 49 19 58 38
22 50 46 14 33 48 25 32 52 69
40 52 57 27 61 42 39 64 52 27

4. The numbers of students attending a class (maximum 25) for 30 lessons are given in the table below:
Students Attending Class
25 24 24 25 24 23
25 24 23 24 25 25
24 25 20 23 25 24
22 24 25 24 23 21
25 23 24 25 24 22

Calculate the range, inter-quartile range and standard deviation using a Frequency Distribution Table.
(This can be found in the answers to the Topic ‘Graphs’).

5. The Shot Put distances thrown by 27 world champion shot putters are given in the table below. The
unit is metres (m).

22.25 20.19 21.39 21.25 21.19 22.07

21.72 20.37 20.45 23.09 21.19 21.22
21.07 21.55 23.12 20.91 22.58 21.97
20.22 20.38 22.37 22.19 21.70 20.54
22.67 21.58 21.72

Calculate the range, inter-quartile range and standard deviation using a grouped Frequency
Distribution Table. (This can be found in the answers to the Topic ‘Graphs’)

Page 48
 Numeracy

Topic 5: Comparing like data

The results of two classes of maths students are compared. This is an example where like data (exam results)
exists in two different settings (classes). It is possible to compare these. To do this we will use Box and
Whisker diagrams.

Before continuing, it is important to make a point about the limitations of this process. When samples of
data are collected there is always random variation between samples. For example; an environmental
scientist collected tadpoles from two different ponds each day for a week. The number of tadpoles in the
first pond was lower than the second pond. It was suspected that the first pond receives water runoff from an
industrial area which is polluted. The lower number of tadpoles in the first pond could be explained by one
of two explanations: (a) chance variation or (b) the first pond was polluted. Perhaps if the difference is
small, the variation is chance variation. If the difference is large, the variation is due to pollution in the first
pond. What is a small or large variation?

In this module it is not possible to separate chance variation from variation due to some effect. In most first
year university statistics courses the topic ‘Hypothesis Testing’ is covered. Hypothesis testing is a process
that gives some certainty to situation outlined above.

Box and Whisker Diagrams

In this section some new data are required. Two brands of batteries were tested on a common battery operated
toy. A sample of twenty batteries of each brand was used. Each battery was used to operate the toy until it
became non-operational. The time taken for this to occur (to the nearest hour) was recorded.

The data collected are listed in the table below:

Brand A Brand B
53 40 37 41 40 56 56 60 50 52 27 18 60 97 35 79 55 73 44 68
58 55 53 52 54 59 59 64 45 42 77 84 93 84 61 78 69 74 55 78

For each brand a 5 number summary is determined. This was mentioned in the section on ‘Interquartile
Range’. Diagrammatically, it is represented as below:

Lowest First Second Third Highest

Value Quartile Quartile or Quartile Value
Median

A five number summary is necessary for drawing a box and whisker diagram.

One way to organise this data is to use a back to back stem and leaf plot.
This method of organising data works well here because the data are easily put into a stem and leaf plot
(see section 2 Tables) and some trends about the data may be seen from the plot.

Page 49
Student Learning Zone | +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone [last edited July 2024]
 Brand A 5th value Stem Brand B
th
6 value 1 8 5th value
6th value
2 7
17th value
7 3 5
52100 4 4 10th value
18th value 998665543322 5 55
40 6 0189 11th value
11th value
7 347889
18th value
10th value 8 44
9 37 17th value
2|4 means 24 hours

The 5 number summary for the Brand A was:

Lowest First Quartile is the Median is the Third Quartile is the Highest
Value = 37 n +1 n +1 3 ( n + 1) Value = 64
4 2 4
20 + 1 20 + 1 3 × 21
= = =
4 2 4
=
21
=
21 = 17.75th value
4 2 This means that the third
= 5.25th value = 10.5th value quartile is three quarters of
This means that the first Median = the way between the 17th and
quartile is a quarter of the way (53+54)÷2=53.5 18th values.
between the 5th and 6th values.
First Quartile Third Quartile
= 42 + 0.25 x (45- = 59+0.75x(59-59)
42) =59
= 42.75
Lowest First Quartile (Q1) = Median = Third Quartile (Q3) Highest
Value = 37 42.75 53.5 = 59 Value = 64

To draw a box and whisker plot:

1) Draw short vertical lines at the 5 values from the 5 number summary.

2) Draw a box from the first quartile to the third quartile.

3) Connect the first quartile to the lowest value with a whisker. Likewise the third quartile to the
highest value. These should be located in the middle between the top and bottom of the
box. You will notice that the median lies somewhere within the box encompassing the
interquartile range.

Page 50
 Five value summary of brand A data used in constructing box and whisker plot:
37 42.75 53.5 59 64

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

The 5 number summary for the Brand B was:

Lowest First Quartile is the Median is the Third Quartile is the Highest
Value = 18 n +1 n +1 3 ( n + 1) Value = 97
4 2 4
20 + 1 20 + 1 3 × 21
= = =
4 2 4
=
21
=
21 = 17.75th value
4 2 This means that the third
= 5.25th value = 10.5th value quartile is three quarters of
This means that the first Median = the way between the 17th and
quartile is a quarter of the way (69+73)÷2=71 18th values.
between the 5th and 6th values.
First Quartile Third Quartile
= 55+ 0.25 x (55- = 84+0.75x(84-84)
55) =84
= 55
Lowest First Quartile (Q1) = Median = Third Quartile (Q3) Highest
Value = 18 55 71 = 84 Value = 97

Drawing the box and whisker plot:

18 55 71 84 97

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Page 51
Now to do the comparing the Box and Whisker Plots should be arranged side by side.

18 55 71 84 97

Brand B

Brand A
37 42.75 53.5 59 64

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Comparing:

1. Based on the lengths of the overall plots and the lengths of the boxes, Brand B batteries are more
variable in lifetime than Brand A.

2. Based on the first quartiles, median and third quartiles, it is possible to state that Brand B will mostly
last longer than Brand A. The lowest value for Brand B is lower than the lowest for Brand A. This is
not a major consideration as lowest or highest values may be ‘odd’ values. Because the lower quartile,
median and upper quartiles are free of ‘odd’ values, they are the most reliable values to base
generalisations on.

Video ‘Comparing Data’

Page 52
Activity

1. The lifetime, in hours, of a sample of 15 light bulbs is: 351, 429, 885, 509, 317, 753, 827, 737, 487,
726, 395, 773, 926, 688, 485.

Construct a box and whisker plot for this data.

2. The Car Speed Data were collected from a school zone just prior to an advertising campaign. The data
collected were:

Car Speed Data (in km/hr)

12 41 44 45 28 40 32 62 46 25
31 35 31 20 59 27 49 19 58 38
22 50 46 14 33 48 25 32 52 69
40 52 57 27 61 42 39 64 52 27

After an advertising campaign to make motorists more aware of speeding in school zones, the
following data were collected.

22 45 42 27 27 40 45 26 75 25
31 48 30 16 28 42 19 55 46 38
34 50 14 59 33 42 30 32 36 18
38 41 50 14 31 39 25 46 39 27

Draw side by side box and whisker plot to determine if the campaign was successful.

3. The heart rate of a group of athletes was compared to the heart rate of a group of office workers after
climbing a set of stairs. The data are presented below:

Athletes
96 111 88 79 101 91
104 106 121 93 103 96
85 117 126 97 83 93
112 106 110 116 91 88

Office Workers
114 107 94 113 113 97
118 103 121 127 117 131
145 132 118 108 145 126
100 138 120

Draw side by side box and whisker plots and make generalisations about the two groups.

Page 53
 Numeracy

Topic 6: Correlation and Regression

In this section, the relationship between two variables is being considered.
Correlation is an attempt to measure the strength of the relationship between the two variables. For
example: a retail store may vary the amount they spend on advertising and then measure the revenue
obtained. There seems to be a natural link between the amount spent on advertising and the revenue
obtained from sales. Calculating the correlation coefficient will give information about the strength and
nature of the linear relationship.

Once the correlation coefficient suggests a reasonable linear relationship exists, the linear relationship can
be described by an equation. The process of finding the “theoretical” or “ideal” linear equation is called
Regression.

The scenario used as an example is relevant to university study. Many studies have shown that time on task
is an important indicator to the overall success of students at university. In this example the success of
students on an examination (as a percentage) will be one variable and the hours spent studying will be the
other variable. The data will be entered directly into Excel. The spreadsheet is shown below:

Page 54
Student Learning Zone | +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone [last edited July 2024]
The next step is to draw the graph of the data. For the scenario, the dependent variable is Exam Result%
because the underlying relationship could be that Exam Result% depends upon the Study Time. Study Time
is the independent variable. When two variables are graphed together, the graph is called a scatter plot or
scatter diagram.
To graph the data from above
>Highlight the data in the table
>Click on the Insert tab
> In the Charts Area click on Scatter
> Choose the option that plots points (not lines)

The correlation can now be considered.

Strong correlation is indicated by the points making a straight line. The straight line may have a positive
slope, indicating positive correlation or a negative slope, indicating negative correlation. If the points form a
line with slope close to zero or seem to be randomly arranged, then there is no correlation.
For positive correlation, it is possible to say 'when the value of one variable increases, the value of the other
variable will also increase'.

For negative correlation, it is possible to say 'when the value of one variable increases, the value of the other
variable will decrease'.

100

90

80
Exam Result%

70

60

50

40

30
0 5 10 15 20 25 30 35
Study Time (Hours)

The appearance of the graph suggests positive correlation. Generally speaking this means that as the Study
Time increases there is an increase in the Exam Result.

Correlation has a number associated with it called the correlation coefficient. The correlation coefficient
ranges from -1 to 1. A correlation coefficient of -1 represents strong negative correlation, 0 represents no
correlation, 1 represents strong positive correlation. The following table shows the link between the
appearance of graphs, the correlation, possible correlation values and the link between the two variables.

Page 55
Graph Correlation Correlation Link between variables
description coefficient
10 The greater the value of
8 the x variable, the
6 greater the value of the y
4
Strong variable.
0.8 to 1
2
Positive
0
0 5 10

10 There is some evidence

8 to suggest that the
6 Quite greater the value of the x
4 Strong 0.6 to 0.8 variable, the greater the
2 Positive value of the y variable.
0
0 5 10
10 There is little evidence to
8 suggest that the greater
6 the value of the x
Weak variable, the greater the
4 0.4 to 0.6
Positive value of the y variable.
2
0
0 5 10
10 There is no evidence to
8 support a linear
6 relationship.
4
No
-0.4 to 0.4
2
relationship
0
0 5 10

10 There is little evidence to

8 suggest that the greater
6 the value of the x
Weak
4 -0.6 to -0.4 variable, the lower the
Negative value of the y variable.
2
0
0 5 10

10 There is some evidence

8 to suggest that the
6 greater the value of the x
Quite
4
variable, the lower the
Strong -0.8 to -0.6
value of the y variable.
2 Negative
0
0 5 10

Page 56
 10 The greater the value of
8 the x variable, the lower
6 the value of the y
4
Strong variable.
-1 to -0.8
2
Negative
0
0 5 10

Looking at the table above and comparing to the graph for our scenario, it is clear that there is some
correlation. The wording ‘Quite Strong Correlation’ is almost suitable, so a term to suggest slightly lower
correlation could be ‘Reasonable Correlation’. The correlation coefficient could be estimated to be about
0.7; overall the correlation could be described as Reasonable Positive Correlation.

Excel can calculate the correlation coefficient using a formula.

The data must be present in a table. Excel can give a graph for the visual assessment of the correlation and
the CORREL function will give the correlation coefficient.

In the cell (say) A18 write the word 'correlation='

In the cell B18, click on the tab Formulas

> click on More Functions
> click on Statistical
Come down the list to CORREL and click.

Enter array 1 by going to the sheet and highlighting the Exam Result values with no heading.

Enter array 2 by going to the sheet and highlighting the Study Time value with no heading. The
spreadsheet should look like the spreadsheet below.

Page 57
The correlation coefficient was 0.892; there is quite strong positive correlation meaning that there is
evidence to suggest that the greater the amount of Study Time, the higher the amount of Exam Result will
be.
The correlation coefficient confirms the relationship between the variables is stronger than the graph
suggests.

As the correlation is strong it is worthwhile determining the regression equation.

In the graph above, Excel has plotted the regression line. The spreadsheet can be modified to include the
slope and y-intercept of the regression line. From this, the equation can be obtained. (You may need to brush
up on Equations of Straight Lines from the Linear Relationships module.)

In cell B20, write the words “Slope =” and in cell B21 write the words “Y-intercept =” The slope is found
by doing:
Click on the cell C20, click on the Formulas tab
> click on More Functions
> click on Statistical
Come down the list to SLOPE and click.

Enter y values (Exam Result%) by going to the sheet and highlighting the exam result values with no
heading.

Enter x values (Study Time) by going to the sheet and highlighting the study time values with no heading.

The y-intercept is found by doing:

Click on the cell C21, click on the Formulas tab
> click on More Functions
> click on Statistical
Come down the list to INTERCEPT and click.

Enter y values (Exam Result%) by going to the sheet and highlighting the exam result values with no
heading.

Page 58
Enter x values (Study Time) by going to the sheet and highlighting the study time values with no
heading.

The equation of the line is:

=R 1.86T + 31.6

where R is the exam result % and T is study time in hours

Note1: In Business text books, the equation may be written as R = 31.6 T+ 1.86
Note2: In some university units, students are required to calculate correlation coefficients and the
regression equation using formulas. Scientific calculators and computer spreadsheets use these formulas
in these calculations.

Video ‘Correlation & Regression’

The purpose of describing the relationship between the variables as an equation is to predict values.

For example: If a student studies for 24 hours, what exam result would be expected?

=R 1.86T + 31.6
R = 1.86 × 24 + 31.6
R = 76.24

The prediction for the exam result is about 76%.

Because of the least-squares method used to calculate regression lines, the value of the y variable (exam
result) can be calculated from the value of the x variable (study time) but not the other way round. This
means that calculating the hours of study required to get a result of (say) 90% should not be attempted. For

Page 59
this to take place, a new regression equation would have to be recalculated giving study time as the
dependent (y) variable and exam result as the independent (x) variable.

In summary,

(a) The strength of the relationship between two variables is called correlation.
(b) If the graph indicates strong positive or negative correlation supported by the correlation coefficient
then determining the regression is worthwhile.
(c) The regression equation can be used to predict the dependent variable from a given independent
variable.
(d) Prediction should only be within the range of independent values that were used to calculate the
regression equation. In this example the regression equation was based on independent variable values
from 5 to 30 hours. Predicting within the range of independent values is called ‘Interpolation’.
(e) Prediction outside the range of independent values is called ‘Extrapolation’. Dependent values found
by extrapolation should be considered unreliable and this practice should be avoided.
(f) The regression equation only applies to this exam with a cohort of student similar to those used to
form the equation.

Page 60
Activity

1. The table below contains information about different types of milks. The information given is the
grams of fat and the energy of the food in kilojoules.

Milk (1 cup) Amount of Fat (g) Energy (kJ)

Skim 0 336
1% Fat 2.4g 420
2% Fat 4.7g 504
Whole 8g 630
Breast milk 10.7g 735
Goat milk 10g 706
Sheep milk 17g 1113

(a) Enter the data into Excel and produce the scatterplot. The hypothesis is ‘the Energy in the milk
depends on the amount of Fat’. Energy is the dependent variable. Comment on the correlation by
viewing your scatterplot.

(b) Modify your spreadsheet to calculate the correlation coefficient. Comment on this.

(c) Modify your spreadsheet to calculate the slope and y-intercept of the regression line.

2. A large company is making note of the amount spent on advertising each month and then comparing
this with the amount of monthly sales. The data are given in the table below:

Month Advertising ($ x1000) Sales ($ x1000)

Jan 2 77
Feb 2.4 79
Mar 6 105
Apr 3.3 110
May 1.5 85
Jun 3 95
Jul 3 110
Aug 3.6 124
Sept 4 136
Oct 2.7 118
Nov 5 127
Dec 6.5 154

(a) Enter the data into Excel and produce the scatterplot. Comment on the correlation by viewing
your scatterplot.

(b) Modify your spreadsheet to calculate the correlation coefficient. Comment on this.

(c) Modify your spreadsheet to calculate the slope and y-intercept of the regression line.
3. Information about climate is given below for towns on roughly the same latitude but varying
longitudes. The longitude is not given but distance inland (east) of the coastal location of Yeppoon is

Page 61
 given. The purpose of the question is to determine if there is a correlation between the climate of a
location and its distance from the sea. (Distance Inland is the independent variable)

Location Distance January Av. July Av. Temp Annual

inland (km) Temp Max°C Min °C Rainfall
Yeppoon 0 29.3 11.8 885
Rockhampton 26 31.9 9.5 796
Walterhall 36 31.4 8 815
Blackwater 193 34.1 6 542
Emerald 266 34.2 6.9 640
Springsure 276 34 6.2 682
Blackall 549 36 6.9 529
Barcaldine 568 35 7.9 500
Longreach 674 37.3 6.8 434
Bedourie 1131 38.4 7.7 262

Data from [Link]

Using Excel, find the correlation coefficient of the Distance Inland vs the three climate data given.
Calculate the regression line as appropriate and make comments about what other information could be
applicable in this situation

Page 62
 Numeracy

Answers to activity questions

Check your skills
Mean Mode Median
1. For the data (right) about the Average Daily Σx 58.3 There is Is the 5th
=
x = = 6.48
Hours of Sunshine, calculate the mean, mode n 9 no mode value.
and median. Median =
6.9
2. For the same data, calculate the range, Range IQR SD
interquartile range and standard deviation.
8.8 - 2.3 = 6.5 Q1=5.95 1.78
Q3=7.3 sample
IQR=1.35 assumed

3. For the Latitude Data, use a 5 number summary to draw a box and whisker plot.
L=2.3 Q1=5.95 M=6.9 Q3=7.3 H=8.8

2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0

4. Use Excel to do a scatterplot of Hours of Sunshine vs Latitude. Latitude is the independent variable.
Daily Hours of Sunshine for 9 Locations
10
9
Av Daily Hours of Sunshine

8
7
6
5
4
3
2
1
0
0 10 20 30 40 50 60
Latitude (°S)

5. Use Excel to determine the correlation and the regression equation.

Page 63
Student Learning Zone +61 2 6626 9262 | learningzone@[Link] | [Link]/learning-zone [last edited July 2024]
 Correlation = - 0.69 Moderate negative correlation
Regression line:
H= −0.11L + 10.2
where L is the latitude
H is the av. daily hours of sunshine

Organising Data – Tables

1. (a) The speed (in km/hr) of 40 cars passing a school at 9am on a school day.

Cumulative
Speed Tally Frequency
Frequency
10 to but less than 20 km/hr ΙΙΙ 3 3
20 to but less than 30 km/hr ΙΙΙΙ ΙΙΙ 8 11
30 to but less than 40 km/hr ΙΙΙΙ ΙΙΙ 8 19
40 to but less than 50 km/hr ΙΙΙΙ ΙΙΙΙ 10 29
50 to but less than 60 km/hr ΙΙΙΙ ΙΙ 7 36
60 to but less than 70 km/hr ΙΙΙΙ 4 40
40

Page 64
 (b) Enter the data into Stem and Leaf Plot.

Stem Leaf
1 2 4 9
2 0 2 5 5 7 7 7 8
3 1 1 2 2 3 5 8 9
4 0 0 1 2 4 5 6 6 8 9
5 0 0 2 2 2 7 8
6 1 2 4 9
4|5 means 45 km/hr

(c) What percentage of cars were doing 40km/hr or more?

Using the Cumulative Frequency Column, we know that 19 motorists were doing less than 40
km/hr. Therefore the number doing 40km/hr or more is 40 (Total) – 19 = 21motorists.
The number doing 40km/hr or more = 21
The fraction doing 40km/hr or more = 21/40
Make this a percentage: 21/40 x 100 = 52.5 or approximately 53%

2. The number of students attending a class (maximum 25) for 30 lessons is given in the table below:

(a) Are the data discrete or continuous? The data are discrete because the number of student can only
be a whole number.

(b) Enter the data into a frequency distribution table (groups not required). Include a relative
frequency and % relative frequency column.

Number attending Tally Frequency Relative % Relative

Frequency Frequency
1
20 Ι 1
30
= 0.0333 3.3%

1
21 Ι 1
30
= 0.0333 3.3%

2
22 ΙΙ 2
30
= 0.0666 6.7%

5
23 ΙΙΙΙ 5
30
= 0.166 16.7%

11
24 ΙΙΙΙ ΙΙΙΙ Ι 11
30
= 0.366 36.7%

10
25 ΙΙΙΙ ΙΙΙΙ 10 = 0.3333 33.3%
30
Total 30 1.00 100%

Page 65
 (c) What proportion of lessons contained 22 students? The relative frequency for 22 students is
0.067 (to 3 decimal places)

(d) What percentage of lessons were fully attended? From the table - 33.3%

3. Systolic Blood Pressures of 35 patients at a Cardiac Clinic

(a) Enter the data into a Stem and Leaf Plot. Use the key: 11|4 means 114.

Stem Leaf
8 8
9 2 5
10 2 7 9
11 0 2 4 5 8
12 0 1 2 2 4 6 7 8
13 1 3 4 4 7 8
14 1 3 4 5 6 9
15 5
16 1
17 5
18 8
11|4 means 114

(b) If hypertension (high blood pressure) is defined by a systolic blood pressure 140 or above, what
percentage of this group are suffering hypertension? With a Stem and Leaf, the original values
are retained, so counting is required. There are 10 out of 35 which is 28.6% (to 1 d.p.)

4. The Shot Put distances thrown by 27 world champion shot putters are given in the table below. The
unit is metres (m).

22.25 20.19 21.39 21.25 21.19 22.07

21.72 20.37 20.45 23.09 21.19 21.22
21.07 21.55 23.12 20.91 22.58 21.97
20.22 20.38 22.37 22.19 21.70 20.54
22.67 21.58 21.72

(a) Enter the data into a Frequency Distribution Table. Your FDT must have at least 5 groups.

Distance Tally Frequency Cumulative % Relative

Frequency Frequency
5
20 to but less than 20.5 ΙΙΙΙ 5 5 × 100 =
18.5%
27
2
20.5 to but less than 21 ΙΙ 2 7 × 100 =7.4%
27
6
21 to but less than 21.5 ΙΙΙΙ Ι 6 13 × 100 =
22.2%
27
6
21.5 to but less than 22 ΙΙΙΙ Ι 6 19 × 100 =
22.2%
27
4
22 to but less than 22.5 ΙΙΙΙ 4 23
27
× 100 =
14.8%

2
22.5 to but less than 23 ΙΙ 2 25
27
× 100 =7.4%

Page 66
 2
23 to but less than 23.5 ΙΙ 2 27
27
× 100 =
7.4%

Total 27 100%

(b) How many have thrown less than 22m? (Use cumulative frequency to answer this) Using the
cumulative frequency column, find the row that contains " but less than 22" - the answer is
given by the cumulative frequency of this row, 19.

(c) What percentage threw 21 to but less than 22m? (Use a % Relative Frequency Column) This is
two groups of the table. Each group has 22.2%, so 44.4% threw 21 to but less than 22m.

Graphs
1. The height of a plant is measured every Friday morning for twelve weeks. Which type of graph would
be best to show the growth of the plant? Line

2. A ‘sound and vision’ shop sells CDs, DVDs, console games and computer games. Which type of
graph would best display the relative sales of the different items sold? Pie

3. A student wishes to compare the number of motorcycle fatalities between the states. To get a more
accurate picture, data are collected for the years 2007, 2008 and 2009. Which type of graph allows
the data to be presented in one graph? Composite Column

4. The number of students attending a class (maximum 25) for 30 lessons is given in the table below:

Column Graph of Student Numbers in Class

12
10
Frequency

8
6
4
2
0
20 21 22 23 24 25
Number Attending Class

Page 67
5. The Shot Put distances thrown by 27 world champion shot putters are given in the Frequency
Distribution Table below. The unit is metres (m).

(a) Construct a frequency histogram for this information. As a second step, put a frequency polygon on
the histogram.

Shot Put Records

7

4
Frequency

0
20 20.5 21 21.5 22 22.5 23 23.5
Distance Thrown (m)

(b) Construct a cumulative frequency histogram and then add an Ogive to the histogram.

Shot Put Records

30

25
Cumulative Frequency

20

15

10

0
20 20.5 21 21.5 22 22.5 23 23.5
Distance Thrown (m)

Measures of Central Tendency

1. The lifetime, in hours, of a sample of 15 light bulbs is: 351, 429, 885, 509, 317, 753, 827, 737, 487,
726, 395, 773, 926, 688, 485. Calculate the mean, mode and median of the values. Do not organise
into a table.
Page 68
 In order to calculate the median, the values must be ordered.
317, 351, 395, 429, 485, 487, 509, 688, 726, 737, 753, 773, 827, 885, 926

Mean Mode Median

X=
∑X No value is repeated – there is no
mode.
As there are 15 values, the position
of the median is:
n The mode would not significant in n + 1 15 + 1 16
317 + 351 + .... + 885 + 926 this example. = = = 8
X= 2 2 2
15 The 8th value is the median.
X = 619.2 Counting through the ordered list,
The mean is approx. 619 hours. The the 8th value is 688. The median is
mean is a good indicator of a central also a good indicator or typical
or typical value in this question. value in this question.

2. The number of children in a 10 families is: 1, 5, 2, 2, 2, 3, 1, 4, 3, 2. Calculate the mean, mode and
median of the values. Do not organise into a table.

In order to calculate the median, the values must be ordered.

1, 1, 2, 2, 2, 2, 3, 3, 4, 5

Mean Mode Median

X=
∑X The mode is 2, it occurs 4 times, or
has a frequency of 4.
As there are 15 values, the position
of the median is:
n The mode is a good indicator of a n + 1 10 + 1 11
1 + 1 + .... + 4 + 5 typical value in this example. = = = 5.5
X= 2 2 2
10 The 5.5th value is the median.
X = 2.5 The 5th value is 2 and the 6th value is
The mean is 2.5 children. The 2 also, so the median is 2. The
mean is a good indicator of a median is also a good indicator or
central or typical value in this typical value in this question.
question. However it is not
possible to have 2.5 children

Page 69
3. For the Car Speed Data, calculate the mean, mode and median of the values after organising the data
in Frequency Distribution Table (This can be found in the answers to the Topic ‘Graphs’).

Car Speed Data (in km/hr)

Cumulative f ×x
Speed(x) Group Midpoint Frequency(f)
Frequency
10 to but less than 20
15 3 3 45
km/hr
20 to but less than 30
25 8 11 200
km/hr
30 to but less than 40
35 8 19 280
km/hr
40 to but less than 50
45 10 29 450
km/hr
50 to but less than 60
55 7 36 385
km/hr
60 to but less than 70
65 4 40 260
km/hr
n =Σf =40 Σfx =
1620

Mean Mode Median

ΣfX The modal group is 40 to but As there are 40 values, the position of the median is:
X= less than 50 km/hr. n + 1 40 + 1 41
Σf = = = 20.5
1620 2 2 2
X= The 20.5th value is the median.
40
The median is the average of the 20th and 21st values.
X = 40.5 The group ’40 to but less than 50 km/hr’ contains the
The mean is approx. 40.5 20th to the 29th value.
km/hr. The mean is a good  n +1
indicator of a central or typical   -cf m −1
Lm + 
2 
value in this question. Median = × Group width
fm
20.5 − 19
Median=40+ × 10
10
Median = 41.5

Page 70
There is also a graphical method for working out the median.

C.F. Ogive for Car Speed Data

45
40
35
30
25
C.F.

20
15
Median is approx. 42
10
5
0
10 20 30 40 50 60 70
Speed (km/hr)

4. Students Attending Class

Number attending Frequency(f) f ×x Cumulative

(x) Frequency
20 1 20 1
21 1 21 2
22 2 44 4
23 5 115 9
24 11 264 20
25 10 250 30
Total Σf =30 Σfx =
714

(a) Calculate the mean, mode and median.

Mean Mode Median

ΣfX The modal group is 24 students. As there are 30 values, the position of
X= the median is:
Σf
n + 1 30 + 1 31
714 = = = 15.5
X= 2 2 2
30 The 15.5th value is the median.
X = 23.8 The median is the average of the 15th
The mean is approx. 23.8 students. and 16th values. The group 24
The mean is a good indicator of a contains the 10th to the 20th values,
central or typical value in this including the 15th and 16th values. The
question. median is 24.

Page 71
 (b) Discuss the appropriateness of each measure of central tendency as a typical value.

All measures of centre are relevant. On average 24 students attend lectures.

5. The Shot Put distances thrown by 27 world champion shot putters are given in the table below. The
unit is metres (m).

Distance (x) Group Frequency Cumulative f ×x

Centre (f) Frequency
20 to but less than 20.5 20.25 5 5 101.25
20.5 to but less than 21 20.75 2 7 41.5
21 to but less than 21.5 21.25 6 13 127.5
21.5 to but less than 22 21.75 6 19 130.5
22 to but less than 22.5 22.25 4 23 89
22.5 to but less than 23 22.75 2 25 45.5
23 to but less than 23.5 23.25 2 27 46.5
Total Σf =27 Σfx =
581.75

Calculate the mean, mode and median.

Mean Mode Median

ΣfX This data are bimodal. The As there are 27 values, the position of the
X= modal groups are 21 to median is:
Σf but less than 21.5 and n + 1 27 + 1 28
581.75 21.5 to but less than 22. = = = 14
X= 2 2 2
27 The 14th value is the median.
The mode is of limited
X = 21.5 use in this example. The median is located in the group ’21.5 to
The mean is approx. 21.5 but less than 22 m’.
m. The mean is a good  n +1
indicator of a central or   -cf m −1
Lm + 
2 
typical value in this Median = × Group width
fm
question.
14 − 13
Median=21.5+ × 0.5
6
Median = 21.6 (rounded to 1dp)

Page 72
There is also a graphical method for working out the median.

Shot Put Records

30

25
Cumulative Frequency

20

15

10
Median is approx 21.6

0
20 20.5 21 21.5 22 22.5 23 23.5
Distance Thrown (m)

Measures of Spread
1. The lifetime, in hours, of a sample of 15 light bulbs is: 351, 429, 885, 509, 317, 753, 827, 737, 487,
726, 395, 773, 926, 688, 485.

In order to calculate the median, the values must be ordered.

317, 351, 395, 429, 485, 487, 509, 688, 726, 737, 753, 773, 827, 885, 926

Range SD
Range = Highest Value - Lowest Value Using a Scientific Calculator
Range = 926 - 317 s = 203 (rounded to nearest whole number)
Range = 609
IQR
The first quartile (Q1) is the The third quartile (Q3) is the
n +1 3 ( n + 1)
4 4
15 + 1 3 × 16
= =
4 4
16 48
= =
4 4
= 4th value = 12th value
Q1 is 429 Q3 is 773
The IQR is 773 – 429 = 344
2. The number of children in a 10 families is: 1, 5, 2, 2, 2, 3, 1, 4, 3, 2. Calculate the range, inter-quartile
range and standard deviation of the values. Do not organise into a table.

In order to calculate the median, the values must be ordered.

Page 73
 1, 1, 2, 2, 2, 2, 3, 3, 4, 5

Range SD
Range = Highest Value - Lowest Value Using a Scientific Calculator
Range = 5 - 1 s = 1.27 (rounded to 2dp)
Range = 4 (assuming sample)
IQR
The first quartile (Q1) is the The third quartile (Q3) is the
n +1 3 ( n + 1)
4 4
10 + 1 3 × 11
= =
4 4
11 33
= =
4 4
= 2.75th value = 8.25th value
Q1 is 1.75 Q3 is 3.25
The IQR is 3.25 – 1.75 = 1.5

3. For the Car Speed Data, calculate the range, inter-quartile range and standard deviation of the values
after organising the data in Frequency Distribution Table. The assumption is that the original values
are lost and it is a sample.

Car Speed Data (in km/hr)

Group Cumulative f ×x
Speed(x) Frequency(f)
Midpoint Frequency
10 to but less than 20
15 3 3 45
km/hr
20 to but less than 30
25 8 11 200
km/hr
30 to but less than 40
35 8 19 280
km/hr
40 to but less than 50
45 10 29 450
km/hr
50 to but less than 60
55 7 36 385
km/hr
60 to but less than 70
65 4 40 260
km/hr
n =Σf =40 Σfx =
1620

Page 74
 Range SD
Range = Highest Value - Lowest Value Using a Scientific Calculator
Range = 70 - 10 s = 14.5 (rounded to 1dp)
Range = 60 (assuming sample)
IQR
The first quartile (Q1) is the The third quartile (Q3) is the
n +1 3 ( n + 1)
4 4
40 + 1 3 × 41
= =
4 4
41 123
= =
4 4
= 10.25th value = 30.75th value
using interpolation using interpolation
10.25 − 3 30.75 − 29
The first quartile (Q1 ) = 20 + ×10 The third quartile (Q3 ) = 50 + ×10
8 7
7.25 1.75
= 20 + ×10 = 50 + ×10
8 7
= 29.1 = 52.5
The IQR is 52.5 – 29.1 = 23.4

The Ogive can also be used to calculate the quartiles and IQR

C.F. Ogive for Car Speed Data

45
40
35
30
25
C.F.

20
15 IQR = 52 – 29 = 23
10
5
0
10 20 30 40 50 60 70
Speed (km/hr)

Page 75
4. The number of students attending a class (maximum 25) for 30 lessons is given in the table below. The
assumption is that the original values are lost and it is a sample.

Number attending Frequency(f) f ×x Cumulative

(x) Frequency
20 1 20 1
21 1 21 2
22 2 44 4
23 5 115 9
24 11 264 20
25 10 250 30
Total Σf =30 Σfx =
714

Range SD
Range = Highest Value - Lowest Value Using a Scientific Calculator
Range = 25 - 20 s = 1.27 (rounded to 2dp)
Range = 5 (assuming sample)
IQR
The first quartile (Q1) is the The third quartile (Q3) is the
n +1 3 ( n + 1)
4 4
30 + 1 3 × 31
= =
4 4
31 93
= =
4 4
= 7.75th value = 23.25th value
The 7th and 8th values are both 23. The 23rd and 24th values are both 25.
Q1 = 23 Q3 = 25
The IQR is 25 – 23 = 2

5. The Shot Put distances thrown by 27 world champion shot putters are given in the table below. The
unit is metres (m). The assumption is that the original values are lost and it is a sample.

Distance (x) Group Frequency Cumulative f ×x

Centre (f) Frequency
20 to but less than 20.5 20.25 5 5 101.25
20.5 to but less than 21 20.75 2 7 41.5
21 to but less than 21.5 21.25 6 13 127.5
21.5 to but less than 22 21.75 6 19 130.5
22 to but less than 22.5 22.25 4 23 89
22.5 to but less than 23 22.75 2 25 45.5
23 to but less than 23.5 23.25 2 27 46.5
Total Σf = 27 Σfx =581.75

Range SD

Page 76
 Range = Highest Value - Lowest Value Using a Scientific Calculator
Range = 23.5 - 20 s = 0.902 (rounded to 3dp)
Range = 3.5 (assuming sample)
IQR
The first quartile (Q1) is the The third quartile (Q3) is the
n +1 3 ( n + 1)
4 4
27 + 1 3 × 28
= =
4 4
28 84
= =
4 4
= 7th value = 21st value
Interpolation is not required here as the By interpolation
7th value is 21. 21 − 19
The third quartile (Q3 ) =
22 + ×10
4
2
= 22 + × 0.5
4
= 22.25
The IQR is 22.25 – 21 = 1.25

The Ogive can also be used to calculate the quartiles and IQR

Shot Put Records

30

25
Cumulative Frequency

20

15
IQR = 22.2 – 21 = 1.2
10

0
20 20.5 21 21.5 22 22.5 23 23.5
Distance Thrown (m)

Page 77
Comparing Data
1. The lifetime, in hours, of a sample of 15 light bulbs is: 351, 429, 885, 509, 317, 753, 827, 737, 487,
726, 395, 773, 926, 688, 485.

In order to calculate the median, the values must be ordered.

317, 351, 395, 429, 485, 487, 509, 688, 726, 737, 753, 773, 827, 885, 926

From the Measures of Central Tendency section, the median is 688.

From the Measures of Spread section, the quartiles are 429 and 773.

The lowest and highest values are 317 and 926.

317 429 688 773 926

300 350 400 450 500 550 600 650 700 750 800 850 900 950

2. The Car Speed Data were collected from a school zone just prior to an advertising campaign. The data
collected were:

Car Speed Data (in km/hr)

12 41 44 45 28 40 32 62 46 25
31 35 31 20 59 27 49 19 58 38
22 50 46 14 33 48 25 32 52 69
40 52 57 27 61 42 39 64 52 27

After an advertising campaign to make motorists more aware of speeding in school zones, the
following data were collected.

22 45 42 27 27 40 45 26 75 25
31 48 30 16 28 42 19 55 46 38
34 50 14 59 33 42 30 32 36 18
38 41 50 14 31 39 25 46 39 27

To organise the data, a back to back stem and leaf plot will be used.
After advertising Stem Prior to Advertising
9 8 6 4 4 1 2 4 9
8 7 7 7 6 5 5 2 2 0 2 5 5 7 7 7 8
9 9 8 8 6 4 3 2 1 10 0 3 1 1 2 2 3 5 8 9
8 6 6 5 5 2 2 2 1 0 4 0 0 1 2 4 5 6 6 8 9
9 50 0 5 0 2 2 2 7 8 9
6 1 2 4 9
5 7

For the Prior to Advertising Data

Page 78
 Lowest First Quartile is the Median is the Third Quartile is the Highest
Value = 12 n +1 n +1 3 ( n + 1) Value = 69
4 2 4
40 + 1 40 + 1 3 × 41
= = =
4 2 4
=
41
=
41 = 30.75th value
4 2 This means that the third
= 10.25th value = 20.5th value quartile is three quarters of
This means that the first Median = 40 the way between the 30th and
quartile is a quarter of the way 31th values.
between the 10th and 11th
values. Third Quartile
First Quartile = 50 + 0.75 x (52-
= 27 + 0.25 x (28- 50)
27) =51.5
= 27.25
Lowest First Quartile (Q1) = Median = Third Quartile (Q3) Highest
Value = 12 27.25 40 = 51.5 Value = 69

For the After the Advertising Data

Lowest First Quartile is the Median is the Third Quartile is the Highest
Value = 14 n +1 n +1 3 ( n + 1) Value = 75
4 2 4
40 + 1 40 + 1 3 × 41
= = =
4 2 4
=
41
=
41 = 30.75th value
4 2 This means that the third
= 10.25th value = 20.5th value quartile is three quarters of
This means that the first Median = the way between the 30th and
quartile is a quarter of the way (34+36)/2=35 31th values.
between the 10th and 11th
values. Third Quartile
First Quartile = 42 + 0.75 x (45-
= 27 42)
=44.25
Lowest First Quartile (Q1) = Median = Third Quartile (Q3) Highest
Value = 14 27 35 = 44.25 Value = 75

Page 79
 12 27.25 40 51.5 69

Prior

After
14 27 35 44.25 75

10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

The quartiles and the median are lower after the campaign, based on this there is some evidence
of a drop in speed. The lowest and highest value have increased, these can be outliers and so
unreliable. The Stem and Leaf Plot suggests that the highest is an ‘odd’ value.

3. The heart rate of a group of athletes was compared to the heart rate of a group of office workers after
climbing a set of stairs. The data are presented below:

Athletes
96 111 88 79 101 91
104 106 121 93 103 96
85 117 126 97 83 93
112 106 110 116 91 88

Office Workers
114 107 94 113 113 97
118 103 121 127 117 131
145 132 118 108 145 126
100 138 120

Office Workers Stem Athletes

7 9
8 3 5 8 8
7 4 9 1 1 3 3 6 6 7
8 7 3 0 10 1 3 4 6 6
8 8 7 4 3 3 11 0 1 2 6 7
7 6 1 0 12 1 6
8 2 1 13
5 5 14

Page 80
For the Athletes
Lowest First Quartile is the Median is the Third Quartile is the Highest
Value = 79 n +1 n +1 3 ( n + 1) Value = 126
4 2 4
24 + 1 24 + 1 3 × 25
= = =
4 2 4
=
25
=
25 = 18.75th value
4 2 This means that the third
= 6.25th value = 12.5th value quartile is three quarters of
This means that the first Median = the way between the 18th and
quartile is a quarter of the way (97+101)/2=99 19th values.
between the 6th and 7th values.
First Quartile Third Quartile
= 91 = 110 + 0.75 x
(111-110)
=110.75
Lowest First Quartile (Q1) = Median = Third Quartile (Q3) Highest
Value = 79 91 99 = 110.75 Value = 126

For the Office Workers

Lowest First Quartile is the Median is the Third Quartile is the Highest
Value = 94 n +1 n +1 3 ( n + 1) Value = 145
4 2 4
21 + 1 21 + 1 3 × 22
= = =
4 2 4
=
22
=
22 = 16.5th value
4 2
= 5.5th value = 11th value Third Quartile
First Quartile Median = 114 = 127 + 0.5 x (131-
= 107 + 0.5 x (108- 127)
107) =129
= 107.5
Lowest First Quartile (Q1) = Median = Third Quartile (Q3) Highest
Value = 94 107.5 114 = 129 Value = 145

79 91 99 110.75 126

Athletes

Office
Workers
94 107.5 114 129 145
70 75 80 85 90 95 100 105 110 115 120 125 130 135 140 145 150

It is quite clear that there is a difference between the two groups.

Correlation and Regression
1. (a) Excel scatterplot. The scatterplot suggests Very Strong Correlation.

Page 81
 Energy Content of various Milks
1200

1000

800

Energy (kJ)
600

400

200

0
0 5 10 15 20
Percentage Level of Fat (%)

(b) The correlation coefficient = 0.989; This also suggests very strong correlation.

(c) The regression line;

=y 44.4 x + 300 or
=E 44.4 F + 300
where E - energy content of the milk
F - percentage Fat level of the milk

2. A large company is making note of the amount spent on advertising each month and then comparing
this with the amount of monthly sales.

(a) The scatterplot suggests reasonable to strong positive correlation.

Chart Title
180
160
140
Sales ($x1000)

120
100
80
60
40
20
0
1 2 3 4 5 6 7
Advertising Cost ($x1000)

(b) The correlation coefficient = 0.735 suggesting Quite Strong Positive Correlation

(c) The regression line.

=y 11.2 x + 70 or
=S 11.2 A + 70
where S - Sales per month ($000)
A - Advertising Costs per month ($000)

Page 82
3. (a) Distance Inland vs January Average Max Temp. The scatterplot suggests strong positive
correlation.
January Max Temp for Towns
40

38

Temperature (°C) 36

34

32

30

28

26
0 200 400 600 800 1000 1200
Inland Distance (km)

Correlation coefficient = 0.92; this suggests Strong Positive Correlation.

Regression line:

=y 0.0071x + 31.5 or
=
TJan max 0.0071d + 31.5
where TJan max - Jan Average Max Temp ( °C)
d - Inland Distance (km)

(b) Distance Inland vs July Average Minimum Temp. There is some (negative?) correlation here. It
appears that once you get 200km inland that the minimum temperature stops decreasing and then
increases at a lower rate.
July Min Temp for Towns
13
12
11
Temperature (°C)

10
9
8
7
6
5
4
0 200 400 600 800 1000 1200
Distance Inland (km)

Correlation coefficient = -0.37. The bounds for "no correlation" are between -0.4 and 0.4, as
the coefficient falls within these bounds, the regression equation is not calculated.

Page 83
(c) Distance Inland vs Annual Rainfall. Scatterplot suggests strong negative correlation. This
means the greater the distance inland, the lower the annual rainfall.

Annual Rainfall for Towns

1000
900

Average Annual Rainfall (mm)

800
700
600
500
400
300
200
100
0
0 200 400 600 800 1000 1200
Distance Inland (km)

Correlation coefficient = -0.94; this suggests Strong Positive Correlation.

Regression line:
y= −0.505 x + 796 or
RAv = −0.505d + 796
where RAv - Average Annual Rainfall (mm)
d - Inland Distance (km)

Page 84
Comments:

Graph (a) is not surprising. Maximum temperature can be influenced by other factors such as altitude but
distance from the sea is a significant one.

Graph (b) is a very interesting graph. Perhaps altitude has an effect over 200 km inland. This could lead to
further investigations.

Graph (c) is expected. However rainfall can be affected by topography such as mountain ranges. This
usually creates a rainy side or a rain shadow depending on the direction of the prevailing wind.
There is one location (Blackwater) that is lower than the regression line suggests, perhaps this
location is in a rain shadow.

Page 85