You are on page 1of 12

INTRODUCTION TO

STATISTICS

Introduction

What is statistics?

Statistics is the science of extracting


information from data. i.e., statistics makes
sense of data.
Statistics is the science of learning from data.
What is the role of statistics in the
research process?

Aspects of statistics
Collection of data
Order, graph and summarize data
(Descriptive statistics)

Refers to graphical and tabular methods used


to summarize and order data

Draw conclusions from data (Statistical


inference)

Refers to methods used to make conclusions


about a population from sample data

Collection of data
To be able to extract information
from data we need data.
This can be a difficult task. The
manner in which the data is
collected will influence the
conclusion being made.
Need to plan how the data should
be collected. (See Chapter 2)

Order, graph and summarize data


Mark Frequency
(Descriptive statistics) Module
0-20
64

Raw data will not necessarily


deliver any meaningful conclusions.
Properly summarized data will make
it easier to determine the
underlying properties of the data .
More detail in chapter 3 to 5

Module Mark
80
60
40
20

Hours studied

0
0-20

Hours
Student studied
20-40 1
2
40-60 3
60-80 4
80-100 5
6
Grand Total
7
8
9
Hours studied
10
11
5-14
12
15-24 13
14
25-34 15
35-44 16
17
45-54 18
19
55-64 20
65-74 21
22
75-84 23
85-94 24
25
95-104 26
27
Grand Total
28
29

20-40

40-60

60-80

80-100 50
40
30
20
10
0

Final module
mark
67
5692.23
82
55.35
6843.90
33
66 2.28
43
22
4664.41
95
30061.79
14
15.66
38
41.00
86
75.59
Frequency
33
51.13
35
73.15
26
55
17.00
49
42.99
34
60
73.26
41
60
37.97
44
26.91
29
9
91.28
31
45
28.66
35
92.42
36
68
60.58
59
32.24
30
36
75.14
32
76
21.72
75
35.93
35
76
57.52
6
60
24.70
48
86.80
300
71
57.48
16
56.60

Draw conclusions from data


(Statistical inference)

It is the method used to draw conclusions from


the data.
In the midst of uncertainty, there must be an
accountable procedure which can be used to
obtain answer from data.

Measurement

Measurement involve the process of assigning a


numerical value to a property of an observation.
A particular persons opinion, for or against, can
be assigned as numerical values, 0 and 1.
Measurements are valid if they lead to useful
information concerning the properties being
studied.
How would you measure success?

Variables

A variable is any property of an observed


element that varies from one element to the
next; it takes on values within a range of valid
values.

Examples:
Height of a person
Weight of a banana
Marks obtained by a student in an exam
gender

Types of variables

Two types of variables appears in statistics,


namely discrete and continuous.

A variable is said to be discrete if the possible


values that it can take on are distinguishable
and disconnected from one another.
A variable is said to be continuous if the possible
values that it can take on are not
distinguishable. i.e., for any two possible values
of the variable it is possible to find another that
lies between them.

Discrete vs. continuous


data

Discrete data: obtained from discrete


variables. 100% accurate measurements
are possible.

Continuous data: obtained from


continuous variables. Values have been
rounded and therefore not 100%
accurate.

Types of scales
Example

Scale

Description

Nominal
scale

The values of the variable only


indicate classes or categories

Colour of a persons eye:


blue, brown, green
0, 1, 2

Ordinal
scale

Is the same as the nominal


scale but has also an order
associated with the categories.

Employees job
satisfaction:
poor, reasonable, good,
excellent
0, 1, 2, 3

Interval
scale

The same as ordinal scale, but


meaning has now been
attached to the differences
between the values. The ratio
has no meaning because the
starting point is arbitrarily
chosen.

Time of day:
14:00-15:30 is the same
as the interval between
8:00-9:30

Ratio scale

The same as interval scale, but Mass and height:


ratios have meaning. The
Starting point of zero
starting point is fixed and
influences the interpretation.
p://psychology.ucdavis.edu/faculty_sites/sommerb/sommerdemo/scaling/levels.h

The computer in statistics

Most programs that analyse data


statistically are compatible with the use on
a network or with personal computers.
The implication of this is that any person
has access to statistical techniques that can
produce answers in a fraction of a second.
SAS, SPSS, CSS statistica,
S-PLUS, STAT1.3.

You might also like