Page 1
©Mazzarol 2016 all rights reserved
Business Research: Principles
and Processes MGMT6791 Workshop 1C: The Nature of Research & Scientific Method
Professor Tim Mazzarol – UWA Business School
UWA Business School DBA Program [email protected] MGMT6791
©Mazzarol 2016 all rights reserved
Page 2
©Mazzarol 2016 all rights reserved
Prediction in scientific research
“Prediction in research fulfils one of the
basic desires of humanity, to discern the
future and know what fate holds. Such
foresight used to involve studying the
stars or looking at the entrails of
animals.”
• Examples of Prediction in Research:
– Medicine • Ali Razi location of hospital in Bagdad.
• Ignaz Semmelweis hand washing in Vienna.
– Physics • Albert Einstein & Stephen Hawking in theoretical physics.
– Astronomy • Urbain Le Verrier & John Couch Adams discovery of
Neptune.
– Archaeology • Harald Schliemann’s discovery of Troy.
Source: Shuttleworth 2008
Page 3
©Mazzarol 2016 all rights reserved
The rise of predictive science
Hypotheses
Predictions Test of
Predictions
Deduction
Observation
Induction
Source: Shuttleworth 2008
Predictive science is being driven
in part by the increasing power of
computers that allow more
complex modelling.
This can be seen in econometric
modelling and climate science.
Governments and the general
public want predictions to help
them plan and develop policies.
Yet predictions can be wrong (e.g.
Global Financial Crisis).
Or highly politically contentious
(e.g. Global Climate Change).
Page 4
©Mazzarol 2016 all rights reserved
Video – Prediction in research
Page 5
©Mazzarol 2016 all rights reserved
Group Exercise
• Read the article by Shuttleworth and watch the “Climate Change” video.
• In groups discuss the following:
– Why has the debate over global warming “become a shambles”?
– Have climate scientists strayed too fare from the “Scientific Method” in reaching their predictions?
– If the majority of scientists believe that climate change is real, why has it been so difficult for them to win the debate?
– What does this tell us about the risks of predictive science?
Page 6
©Mazzarol 2016 all rights reserved
Conceptual variables
• “Conceptual variables are often expressed in general, theoretical, qualitative, or subjective terms and are important in the hypothesis building process.”
• The research hypothesis “H¹” is usually developed from the conceptual variables.
• To measure conceptual variables: – You need an objective definition.
– Validated measurement instruments.
– Theoretically supported operational variables.
– A consensus for all the above.
Source: Shuttleworth 2008
Page 7
©Mazzarol 2016 all rights reserved
Example of conceptual variable
measurement
• Step 1 – Conceptual variable:
– The effect of a nicotine patch is poorer
among people who lack the mental
determination to quit smoking.
• Step 2 – Define what is meant by:
– “effect of a nicotine patch”, and “mental
determination”.
• Step 3 – Decide on measurement
scale:
– Effect of nicotine patch • (nominal) “Yes” or “No” .
• (ordinal) “none/moderate/high”.
• Based on potency of patch.
– Mental determination • (nominal) “Present” or “Absent” .
• (ordinal) “none/moderate/high”.
Source: Hani 2009
Page 8
©Mazzarol 2016 all rights reserved
Fuzzy concepts to operationalisation
• Operationalization starts with “fuzzy
concepts” (conceptual variables)
and sets out exact definitions for
each variable, increasing the quality
of the results, and improving the
robustness of the research design.
• Fuzzy Concepts
– Vague conceptual ideas that
lack clarity or require validation.
– Require clear definition to
enable accurate replication in
the research process.
Source: Shuttleworth 2008
Page 9
©Mazzarol 2016 all rights reserved
Real and nominal concepts
Concepts can be classified into two types: Real and Nominal
“Real Concept”
Describes the concept’s
characteristics of the
concept.
• Golf club
• Woman
• Golf Ball
• Golf course
Changing the name of the
concept will not change
the characteristics.
Well understood and
observable.
“Nominal Concept”
Has no definitive nature.
Both the name and its
characteristics are
arbitrary.
• Democracy
• Justice
• Peace
• Love
• Tyranny
• Mercy
Subject to value systems
and bias. Need careful
definition.
Page 10
©Mazzarol 2016 all rights reserved
Higher and lower order concepts
Concepts can be classified into a hierarchy higher and lower order
• Animals
• Terrestrial
• Mammals
• Quadrupeds
• Dogs
• Domesticated Dogs
• Sheep Dogs
• Kelpies
• Red Kelpies
High Order
Low Order
A Sheep Dog
Concept Hierarchy
• Animals
• Terrestrial
• Reptiles
• Limbless
• Venomous
• Diurnal
• Serpent
• Dugite
• Coastal Dugite
A Dugite Snake
Concept Hierarchy
Page 11
©Mazzarol 2016 all rights reserved
Group Exercise
• In groups discuss the following:
– Problem 1: Defining real and nominal concepts
– Problem 2: Concept identification
– Problem 3: Higher and lower order concepts
Page 12
©Mazzarol 2016 all rights reserved
Operationalization
Theory
Hypotheses
Observations of the world
Empirical Generalisations
Deduction
Operationalization
Induction
Source: Shuttleworth 2008
Operationalization is the process
of strictly defining variables into
measurable factors.
The process defines “fuzzy”
concepts and allows them to be
measured, empirically and
quantitatively.
Analysis
Page 13
©Mazzarol 2016 all rights reserved
Importance of operationalization
• Case of the Mars Climate Orbiter
– 1998 NASA sent the Mars Climate Orbiter to
space.
– Mission – to study the climate of Mars, the
atmosphere and planet surface.
– 1999 all communication with the space craft was
lost as it entered Mars orbit.
– Cause of failure – mismatched ground based
software that used empirical measures rather that
metric measures. Put the spaceship into the
wrong angle for orbit leading to it entering the
atmosphere and burning up.
– Lessons:
• Good operationalization requires clear
definition and agreement over use of the right
measurement methods for a study.
• This allows all researchers to adopt the same
standards and follow a similar methodology.
• Failure to identify the common units of
measure resulted in a disaster.
Source: Shuttleworth 2008
Page 14
©Mazzarol 2016 all rights reserved
The operationalization process
Source: Experiment-Resources.com
Questions about
the big picture Population
Hypothetical
Theories/Concepts
Theory Literature
Review &
Scientist’s
Thoughts
Defining a
Research
Problem
Sampling Choosing
Indicators
Creating
Study
Testable, narrow
Hypothesis (Prediction)
Sample Observable,
Measurable Variables
Specific Situation Da
ta
Co
lle
cti
on
Ch
oo
sin
g
De
sig
n &
Me
tho
d
Co
nc
ep
tua
l
Va
ria
ble
s
Page 15
©Mazzarol 2016 all rights reserved
Different research methods
• Experimental Research Methods – Common in physical and life sciences as well
as psychology.
– Typically quantitative using interval or ratio scales.
– Usually has experimental and control groups.
– Often viewed as “true science”.
– Can by difficult to do with need for rigorous design, equipment and labs.
– Need to control variables is a key challenge.
– Requires replication and falsification.
• Opinion Based Research Methods – Common in social science.
– Can be qualitative and quantitative with ordinal or interval scales.
– Less expensive than experimental design.
– Less precise than the experimental design.
– Can involve “experimental” type studies that allow for control groups.
– Allows for replication and falsification.
Source: Shuttleworth 2008
Page 16
©Mazzarol 2016 all rights reserved
Video – Experimental design
Page 17
©Mazzarol 2016 all rights reserved
Different research methods
• Observational Research Methods
– Common in social science but also used in behavioural science, anthropology.
– Often dismissed as not robust and “quasi-experimental” in nature.
– Use nominal and ordinal scales plus researcher coding of qualitative data.
– Examples:
• Case Studies
• Focus Groups
• Interviews
• Grounded Studies
– Difficult to replicate or falsify.
– Useful to explore phenomena prior to more rigorous methodology.
– Commonly used in business research to help develop theory.
– Major problem is identifying cause and effect.
Source: Shuttleworth 2008
Page 18
©Mazzarol 2016 all rights reserved
Video – Cause and effect
Page 19
©Mazzarol 2016 all rights reserved
Measurement in research
• Nominal Measures
– Numbers arbitrarily assigned to variables to enable easier manipulation.
– Example:
• Numbering focus groups = “In group 1 we found...”
– Useful only as a point of reference or labelling.
• Ordinal Measures
– Numbering system that has meaning and can be statistically analysed.
– Example:
• Likert scales “1=strongly disagree...5=strongly agree”
– Allows numerical analysis but is lacking in precision.
• Interval Measures
– Measures that have scales with an arbitrary zero point.
– Example:
• Celsius or Fahrenheit temperature scales.
– Allow more precision and scales can be divisible.
– However, scales can have negative measures (e.g. -5ºC).
• Ratio Measures
– Measures that have true zero points and no negative values.
– Example:
• Weight in kilograms.
– Allow most precision and have no relationship of scale,
• e.g. 100 KG being twice as heavy as 50 KG.
Source: Shuttleworth 2008
Page 20
©Mazzarol 2016 all rights reserved
The scientific method
Develop research questions
Review previous research evidence (literature review)
Construct hypotheses
Test hypotheses by experiment or observation
Analyse data and draw conclusions
Were the hypotheses
correct?
The “Scientific Method” is a
standard process for how to
undertake research that aims
to discover new knowledge.
Page 21
©Mazzarol 2016 all rights reserved
Scientific observation
• Scientific observation is the central element of scientific method or process. The core skill of a scientist is to make observations.
• Observations: – Any information, data or knowledge from the outside world
received via our senses, or scientific instruments.
– A description of what you see.
• Inferences: – Assumptions made by us from the observations we make.
– Care must be taken in making inferences as you may not interpret the data correctly.
• Types of observations: – Qualitative – subjective and less precise.
– Quantitative – uses numbers and is objective.
Source: Shuttleworth 2008
Page 22
©Mazzarol 2016 all rights reserved
Video – Observation and inference
Page 23
©Mazzarol 2016 all rights reserved
Group Exercise
• In groups:
– Read the exercise.
– Address the questions.
– What does this tell you about observation and inference?
Page 24
©Mazzarol 2016 all rights reserved
End of presentation