Statistical literacy

Dr Kazunga C
OUTCOMES
• This unit will enable student to:
• Demonstrate an understanding of statistical literacy
• Describe the concepts of statistics
• Critically analyse everyday statistical examples
What is statistics?
• Statistics involves the collection and analysis of data to
describe the world.

• Data is information or facts you find out or are given.

• Data can be word or numbers or a combination of both.

 Four stages in data analysis

Freil and Bright (1998), describes four stages in data analysis:

• Ask a question
• Collect data
• Analyse data
• Form and communicate conclusions.

• Although the process appears to be linear, as data is collected

and analysed, researchers may discover unexpected results and
change the questions.
 What thinking characterises statistical literacy?
• Just as people should know how to read and make sense of
words, so they should be able to read and make sense of
numbers.
• In particular here, I am talking about the numbers and graphs
that describe a whole lot of information.
• A statistically literate person uses statistical thinking to deal with
data to gain insight into situations, and to draw correct and valid
conclusions.
• Moore (2000) suggests the following principles of statistical
thinking which show us what it means to be statistically literate.
• I will provide an example to illustrate each principle – see if you
 Data Illuminate
• People often hold opinions about issues that are based on
hearsay, biased opinions or simply informed only by their own
experience.
• We might, for example suspect that women are not promoted
to principal posts in schools as often as men.

• Do we know the facts? Knowing the actual data might throw

light on our ideas and help us to develop informed opinions.

• A statistically literate person recognises the need for data

and seeks out the facts before forming an opinion.
 Data beat Anecdotes
• An anecdote is a short entertaining account of something that
has happened, usually to the person that is telling it.
• We can easily see that such stories will be remembered and
given far more importance in our minds than they really deserve!
• We will remember a woman passed over for promotion telling
her story in a passionate way and tend to give more importance
to her story than to the data about women in high positions.

• A statistically literate person knows that data are more

reliable than anecdotes because they carefully describe an
overall picture rather than focussing on a few incidents.
 Beware the Hidden Variable
• Almost all relationships between two variables are influenced by
other variables hiding in the background.

• For example, we might see data reported that gives the pass/fail
rates for Mathematics at a variety of schools.

• We cannot simply draw conclusions about which school is better

without looking a bit deeper.
 contd
• Perhaps one school has difficult entrance tests and so only
academically successful students are enrolled
• – the results could be less impressive than those for a school
that has an open policy and still manages a good pass rate.
• Including the data on the Mathematical achievement of learners
entering the school might well influence the conclusions we
come to.

• A statistically literate person will look behind the data

before making conclusions.
 Where the Data come from is important

• We often heard people say that the answer to a question

depends on who you ask!
• For example, you cannot expect to get a fair idea of the most
popular football team in Kwazulu Natal by asking people from
one city to name their favourite team.

• A statistically literate person checks that the data is

obtained from a representative sample of people before
drawing conclusions.
 Variation is everywhere
Anxious mothers often weigh their babies every week to check that
they are putting on weight and growing steadily.
Even a healthy baby has weeks when the weight gain is small –
the mother must look at the general trend and not worry too much
about short term variations.

A statistically literate person understands that there will

usually be short term variations in data even when a strong
overall trend is apparent.
 Conclusions are not certain
• We can never be completely certain that the conclusions
reached will apply in a particular case.
• For example, a study might show that the mean height of six
year olds in a school is 1,1m.
• Although we can expect that in most cases the mean height will
be close to this, the variation that is everywhere could easily
lead to a school where the mean height is far greater.
• We can only predict the pattern for a large number of cases.

• A statistically literate person should not be surprised if a

particular case does not fit the general conclusion.
 In summary,

• Statistical thinking is a way of thinking that characterises a

statistically literate person.

• It enables us make sensible and informed decisions based on

the data available to us, which in turn will help us to be better
citizens in a democracy
Basic terminology
• A set of data contains information about a group of
individuals.
• These individuals may be people, or countries, or
organisations etc.
• Each piece of information characteristic of an individual is
known as a variable.
• The data may contain just one piece of information about
each individual (univariate data), two pieces of information
(bivariate data) or many pieces of information (multivariate
data).
 Types of variables
• We can distinguish two types of variables:
• Categorical variables: (nominal variables) These place an
individual into a particular category such as gender, nationality
etc.
• Quantitative variables: These have numerical values for which
arithmetic operations make sense. Examples are typically scores
on tests or measured data.
It is important that you are able to distinguish between the two
types of data as the graphs used to display data, and the statistics
used to describe data differ depending on the type of variable
under consideration.
Collecting data
• Let us collect data to use as a class: How many siblings
you have, age, gender, shoe size, tongue rolling etc?
 Categorical Variables:
• Categorical variables can be displayed in three ways –
pictographs, pie graphs, and bar graphs.
• The values of a categorical variable are the labels for the
category e.g. “male” or “female”.
• The distribution of a categorical variable lists the categories and
gives either the count (frequency of the variable) or percent
(relative frequency of the variable) of individuals who fall into
each category.
• This information is usually tabulated (put into a table) and then
the graph is drawn.
• A tally table is useful to do the count efficiently in large data sets.
 Tally Tables
• This is a way of organising the raw data.

• Firstly we set up the categories and then make a tally mark

each time the data fits into that category.

• A clever idea is to bundle the tally marks into groups of five

which makes the counting at the end much easier.

• We will use the “sibling data” to see exactly how this works.
Tally table

Number Tally Frequency

of siblings
0
1
2
3
4
5
6
7
8
Bar Graphs:
• These graphs are generally easy to understand and used
frequently in the media.
• The number of items in each category is represented by a
bar – the scale tells us what frequency is being shown by
the height of the bar.
• We will use the “sibling data” again draw a graph together
in class to emphasise the important points.
• It is easiest to draw these graphs on squared paper.
Squared paper
Points to Remember:
• The graph must have a fully descriptive heading (as
always!)
• The horizontal axis identifies both the general criterion
for the categories (say gender, or favourite subject) and
labels for the categories, which are written centrally
underneath each bar.
• The vertical axis gives the count or percentage of
individuals falling into each category. The scale must be
clearly marked.
• Bars should be of equal width, and evenly spaced (not
usually touching). The width of the bar has no
Activity 2 page 105

Questions

Questions
Reference

Favourite

subject

Fraction
subject
Gender
number

score

16)
on

Data
Age

(out
of
Total
Best

test
1 13 F HSS Math s 15 8 7
2 13 F English Math s 15 7 8
3 13 F Math s Math s 14 7 7
4 12 F NS NS 12 7 5
5 13 F NS Afrikaans 13 8 5
6 13 F Math s Math s 14 8 6
7 13 F Math s Zulu 12 7 5
8 13 F Math s NS 12 6 6
9 13 F Math s NS 11 7 4
10 13 F NS Math s 11 7 4
11 12 F Math s HSS 8 6 2
12 13 F HSS NS 10 7 3
13 13 F Math s NS 6 3 3
14 12 F NS NS 8 4 4
15 13 F Math s Math s 13 7 6
16 13 F NS NS 11 8 3
17 13 F NS HSS 14 8 6
18 12 F English Math s 12 6 6
19 14 M Math s Math s 13 6 7
20 12 M Math s Math s 14 7 7
21 12 M NS Math s 15 8 7
22 12 M Math s Math s 16 8 8
23 12 M Math s Math s 14 8 6
24 13 M NS Math s 14 8 6
25 13 M HSS Math s 7 4 3
26 13 M English NS 12 8 4
27 13 M NS NS 12 6 6
28 13 M Math s NS 16 8 8
29 12 M Math s Math s 11 6 5
30 13 M LO NS 10 5 5
31 13 M NS Math s 14 7 7
32 13 M Math s NS 13 7 6
33 14 F English English 8 4 4
34 14 F English Science 14 7 7
35 14 F Drama Afrikaans 11 5 6
36 14 F Home Ec BE 15 8 7
37 13 F English English 15 8 7
38 13 F Drama English 12 7 5
SURVEY DATA ABOVE
• 1.Identify the categories in the “Favourite subject” data.
• 2.Draw a tally table to count the number of individuals in
each category(frequency)
• 3.Express this number as a percentage of the
sample(relative frequency)
• 4.Display this data on a bar graph.
 Part one
Categories identified Tally Frequency
C T F R Relative
a a r e
t
e
l
l
e
q
l
a
Frequency
g y u t
o e i
r n v
i c e
e y
s F
r
i e
d q
e u
n e
t n
i c
f y
i
e
d
Bar Graph
END OF LECTURE
• SIYABONGA
• THANK YOU
• DANKIE
• TATENDA