Introduction to Quantitative Research, Questionnaire design and Basic statistical analysis

Dr Andrew Boocock
 • In this lecture we shall:

• Apply the philosophy of positivism

Learnin • Apply questionnaire design principles to a

research question
g • Design questionnaire questions in line
with research question/objectives
Outcom • Determine an appropriate sampling
strategy
es • Plan to collect nominal, ordinal, interval
and ratio data within a questionnaire?
• Plan to analyse quantitative data – An
introduction.
The Research Design

Clear objectives derived from the

research question

Research philosophy outlined, justified

and critiqued

Research instruments designed and

justified

Discussion of Validity and Reliability

Consideration of constraints and ethical

issues impacting on research
The Importance of Research
Philosophy

The research philosophy you adopt has important

assumptions, such as:
• The way in which you view the world around you
• These assumptions will underpin your research strategy and
data collection methods

(Saunders and Lewis 2012:pp.104)

 Positivism versus Interpretivism

Objectivism Subjectivism

Saunders et al. (2005)

Differing Approaches to Research:
Deduction and Induction
Time Horizons: Select the appropriate time horizon

Cross-sectional studies

Longitudinal studies

Saunders et al. (2007)

 Qualitative research –
Interviews, Action research,
focus groups, ethnographic
research, participant
observation, grounded theory
Data
Collectio
n Quantitative research –
Controlled experiments,
Surveys, statistical analysis
Multiple research methods

Reasons for using mixed method designs:

• Triangulation
• Complementarily
• Aid interpretation
• Study different aspects
Qualitative research –
Reminder from previous
video workshop

• Interpretivism
• Subjectivism
• Particularisation rather than generalisation
• Induction
• Longitudinal rather than cross sectional
• Purposive sampling
• Qualitative research – Thematic analysis
• Interviews – validity and reliability?
• Can be used as part of mixed methods approach to
research
Quantitative research

• Positivism – philosophy of objectivism

• Deduction as with scientific research
• Survey research
• Mono-method
• Cross sectional
• Random sampling
• Quantitative research – Descriptive and
inferential statistics
• Can be used as part of mixed methods approach
to research
Surveys

• Key features
• formally designed schedule of questions
• involve a larger representative sample
• rely on information from respondents
• provide quantitative data
• transparency
 Validity
and
Reliability
in • Content validity - instrument adequately covers all the
content that it should
Quantitat • Construct validity -can draw inferences from
ive quantitative data related to the concept being studied.
- Homogeneity (measures one construct), Convergence,
Research Theory evidence (eg behaviour), predictive Validity

• Reliability relates to the consistency of a measure.

Collecting primary data

• Define the research population

• Is there a suitable sampling frame?
• Choose a suitable sample design.
• Choose a suitable data collection method.
• What sample size is required to achieve the required level of
accuracy?
The need to sample

Sampling- a valid alternative to a census when

• A survey of the entire population is impracticable

• Budget constraints restrict data collection

• Time constraints restrict data collection

• Results from data collection are needed quickly

 Types of sampling

Probability (random)

• preferred option in quantitative research

because minimises biases and maximises
generalisability

Non-probability (non-random)

• more suited to qualitative research

Sampling • If random sampling is not undertaken
it is not possible to make
generalisations about the whole
population
 Selecting samples
Population, sample and individual cases

Source: Saunders et al. (2009)

Overview of sampling techniques

Source: Saundes et al. (2009)

Types of Random Sampling
 The • Survey – Involves the structured
collection of data from a sizeable
Survey population. Data collection may take
the form of questionnaires or
structured interviews.
 Types of Questions in a
Questionnaire
Type Use when... Examples
Open ...unsure of the response, If there are any other areas or issues that concern
require a detailed answer, you, please feel free to comment below*
or want to find out what is
uppermost in respondent’s What do you like most about visiting this theme
mind park?

List ...You need to be sure the Which of the following fruit juices have you
respondent has considered purchased in the past week?
all possible answers. (Please tick
all appropriate boxes)
Apple
Orange
Tomato
Pineapple
Mango

(Saunders and Lewis 2012:143)

 Types
Type
of Questions in a
Use when... Examples
Category Questionnaire
...You want the respondent
to place a list in rank order.
Number each of the holiday destinations listed below in order if their
attractiveness to you for your next holiday. Number the most attractive holiday
destination 1, the next 2, and so on. If a destination is not attractive, leave blank.

Destination Attractiveness
England
Germany
Spain
Holland
Wales

Rating ...You want a respondent’s I generally believe what my manager tells me

opinion or belief.
Agree Tend to agree Tend to disagree Disagree

How Likely do you believe it is that you will pay off your overdraft within a year of
graduation?

Very likely Quite likely Not sure Quite Unlikely Very unlikely

Types of Questions in a Questionnaire

 Types
Type of Questions in a Examples
Use when...

Questionnaire
Quantity ...You want the
respondent to tell you
How many dependent children do you have living with you?

a number or amount

(For example -write )

2

Matrix ...the responses to two or The following statements ask about your feelings regarding the future of the
more questions are Happy Toy Manufacturing Company.
selected from the same set
of possible answers. Strongly agree Agree Disagree
Strongly disagree

I feel the future

For the company
Is getting brighter

I would be happy
to spend the rest
of my career at the
company

(Saunders and Lewis 2012:143)

Scaling
Converting a continuum into discrete values or a score
e.g. Likert scale - a statement and a range of responses:

“People watch too much television”

strongly disagree undecided agree strongly

disagree agree
1 2 3 4 5
Scaling
Ranking scale:

Ranking of five most important factors in

employment, from 1 (low) to 5 (high):

high wages ( )
job satisfaction ( )
job rotation ( )
shorter working hours ( )
recognition ( )
Levels of variable measurement

Four levels of data

• Nominal
• name only e.g. countries
• Ordinal
• ranking e.g. cold, warm, hot
• Interval
• equidistant values e.g. satisfaction 1-2-3-4-5
• Ratio
• actual count with a real “zero point”
e.g. # of children, annual salary
Car: The Survey

• Nominal What make of car do you own: ___________

• Ordinal What size is your car?
small/medium/large
• Interval How many kilometres do you drive per year
• 0 – 20 000
• 20 000 – 40 000
• 40 000 – 60 000
• 60 000 – 80 000
• 80 000 +
• Ratio
How much did it cost: ___________
Question research problems (questions)
naire information needs
Design questions linked to research problems
draft design  pilot final design
Process
Questionnaire Design

• Closed (pre-coded) questions

- checklist, ranking, Likert scales

• Open-ended questions
• plain language (no jargon)
• no ambiguity
• one question at a time
• no leading questions
Questionnaire Design

• Ordering of questions
straightforward sensitive
• Cover letter or introductory remarks
- aims/objectives/reasons
- benefits/importance
- request for participation
- confidentiality statement
 Questionnaire
Organization…

The Location of
Questions on the
Questionnaire
Should be Logical
Questionnaire organisation …
Pre-coding questionnaire
• Pre-coding questions refers to the placement of numbers on the
question responses to facilitate data entry and analysis of the
completed survey

• Normally coded as 1 or 2 and every response category is coded

 Questionnaire Organization…
Pre-coding questionnaire
Marital Status
1 1. □ not married A

2
2. □ married

Pre-coding Question #
Record Format Coding

Questionnaire Organization…
Pre-coding questionnaire
Respondent Number 89 A
Marital Status 1 x Not Married B
2 Married
Sex of Respondent 1 x Male C
2 Female
Age of Respondent 28 Years D
Amount Spend $ 14 .00 E
Rating One 2 F
Rating Two 3 G
Rating Three 5 H

Pre-coding
question
 Spreadsheet Data File Format
Questionnaire Organization…
Pre-coding questionnaire

Respondent Number 89
A
Marital Status 1 x Not
Married B
2 Married
Sex of Respondent 1 x Male
C
2 Female
Age of Respondent 28 Years
D
Amount Spend $ 14 .00
E
Rating One 2
F
Data analysis

‘All meanings, we
know, depend on
-George Eliot
the key of
interpretation.’
Presenting and Interpreting Quantitative Data

Presenting quantitative data often involves the production of

graphs and tables
These need to be
• Chosen to make relevant arguments
• informative yet simple, so that they aid reader’s understanding
• All information retrieved in tables/graphs needs to be
explained/discussed
Exploring and presenting data

• Exploratory (descriptive) analysis can include:

• Specific values (frequencies) for each category

• Highest and lowest values
• Trends over time
• Proportions
• Distributions (i.e. how widely dispersed)

• Whatever you choose should be linked to the research question and

objectives
Presenting your statistics
Histogram
Mode, Median and Mean
Displaying Likert Rating
scale
Presenting and Interpreting Quantitative
Data – Inferential statistics

• Associating variables
• How is variable A related to variable B?
• Correlation (Chi Square, Pearson):
• –presence / absence
• –direction
• –strength
Types of variables

Dependent = the variable of primary

interest; it is the main variable that lends
itself for investigation as a viable factor.

Independent = one that influences the

dependent variable in either a positive or
negative way.
How to Interpret Correlation Coefficients
 Examples of theoretical frameworks

Research shows that successful new product

development has an influence on the stock market price;
i.e. the more successful the new product, the higher will
be the stock market price

Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E

Cross-cultural research indicates that managerial values
govern the power distance between superiors and
subordinates.

Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E

Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E THEORETICAL FRAMEWORK ANSWERS TO EXERCISES (PAGES 113-120 OF MANUAL) 5J