business research: principles and processes mgmt6795 business... · correlational designs...

42
©Mazzarol 2017 all rights reserved Business Research: Principles and Processes MGMT6795 Workshop 6B: Preparing the thesis proposal and defence Professor Tim Mazzarol – UWA Business School UWA Business School DBA Program [email protected] MGMT6794 ©Mazzarol 2017 all rights reserved

Upload: haanh

Post on 11-Aug-2018

215 views

Category:

Documents


0 download

TRANSCRIPT

©Mazzarol 2017 all rights reserved

Business Research: Principles and Processes MGMT6795Workshop 6B: Preparing the thesis proposal and defence

Professor Tim Mazzarol – UWA Business School

UWA Business School DBA Program [email protected]

©Mazzarol 2017 all rights reserved

©Mazzarol 2017 all rights reserved

Methodology issues

• Qualitative or Quantitative?• Mixed methodology?• Triangulation• Validity

– Internal – External– Construct– Statistical conclusion

• Reliability

©Mazzarol 2017 all rights reserved

Summary of research methods

Design Type Aims Research Type

Descriptive Designs Observe and Describe • Descriptive research• Case studies• Naturalistic observation• Surveys

Correlational Designs Prediction • Case control study• Observational study• Cohort study• Longitudinal study• Cross-sectional study• Correlational studies in general

Semi-Experimental Designs Determine causes • Field experiment• Quasi-experimental design• Twin studies

Experimental Designs Determine causes • True experimental design• Double-blind experiment

Reviewing other research Explain • Literature review• Meta-analysis• Systematic reviews

Pilot studies Does the proposed design work?

• Pilot tests/studies

©Mazzarol 2017 all rights reserved

Qualitative and Quantitative Research

Theory

Hypotheses

Observations of the world

Empirical Generalisations

Deduction

Operationalization

Induction

Source: Shuttleworth (2008)

Analysis

Qualitative Research Quantitative Research

Mixed-MethodResearch

©Mazzarol 2017 all rights reserved

Quantitative Research Design

Source: Shuttleworth (2008)

Advantages• Can finalise results by testing a

hypothesis with empirical data.

• Is consistent with “the scientific method”.

• Allows for a “defensible” set of findings that can be more easily replicated.

• Is typically viewed as more objective and unbiased.

• Often easier to control variables.

• Offers a useful way to validate theories developed via qualitative studies.

Disadvantages• Can be difficult and expensive to

undertake if an experiment is required.

• Requires the collection of large data sets.

• Requires use of statistical analysis, usually multivariate.

• Care must be taken with sampling, questionnaire design and statistical method.

• Only highlights support or refutation of the null hypothesis.

• May not fully explain why or how.

©Mazzarol 2017 all rights reserved

Qualitative Research Design

Source: Shuttleworth (2008)

Advantages• Valuable when a subject is highly

complex and cannot be easily developed into testable hypotheses.

• Can be undertaken without the need for large scale sampling or the cost of experiments.

• Does not require knowledge of complex statistical methods.

• No need for the development of validated test methods

– (e.g. questionnaire scales).

Disadvantages• Can be very time consuming with the

need to arrange interviews or focus groups, plus the need to transcribe data.

• Data cannot analysed using the same approaches of statistical validation, so some questions may linger over the reliability of the findings.

• Findings offer more a “general guide” as to cause-effect.

• Vulnerable to researcher bias and personal opinion.

• Difficult to replicate.

©Mazzarol 2017 all rights reserved

Quantitative vs. Qualitative Research Design

Source: Shuttleworth (2008)

Goals of Research:

• Qualitative Research – to provide a complete, detailed description of the research topic. Usually more exploratory in nature.

• Quantitative Research – to provide measurable and statistically significant tests of the null hypothesis via statistical models to offer validation of theories.

Qualitative Research Quantitative ResearchHypothesis Broad NarrowDescription Whole picture FocusedType of research Exploration ValidationPhase of study Early Late

©Mazzarol 2017 all rights reserved

Research Method in Management

Source: Scandura & Williams (2000)

ExternalValidity

ConstructValidity

StatisticalConclusion

Validity

InternalValidityTriangulation

Research Strategy

• Formal theory• Literature review• Sample survey• Lab experiment• Simulation

- laboratory- field

• Field study:- primary data- secondary data

• Field experiment• Judgement task

Related Areas:• dependent

variables• substantive

domain

• Time frame• Research

strategy

• Type of sample• Occupation of

subjects• Research

strategy

Construct Validity

• Confirmatory factor analysis

• Exploratory factor analysis

Discriminant/ Convergent/Predictive validity

Interrater reliability

Related Areas• type of dependent

variables• number of data

sources

• Sample size• Number of

dependent variables

• Data analyticapproaches

©Mazzarol 2017 all rights reserved

Triangulation

Source: Scandura & Williams (2000)

• Triangulation:– Using multiple reference points to

locate an object’s exact position• Involves use of different data

collection methods– Surveys– Interviews– Documentary data

• Can involve different research designs– Mixed methodologies– Case studies & surveys– Experiments & interviews

• Should increase validity

©Mazzarol 2017 all rights reserved

Validity issues

Source: Scandura & Williams (2000)

• Internal Validity– Relates to causation– Is there a true cause-effect relationship

between variables?• External Validity

– Can you generalise the findings?– Influenced by research design &

sample selection• Construct Validity

– How well do the measures employed fit the theories & questions?

• Statistical Conclusion Validity– Ability to draw conclusions based on

the results produced

©Mazzarol 2017 all rights reserved

Validity and Reliability

Reliable Not Valid Valid Not Reliable Not Reliable or Valid Reliable and Valid

©Mazzarol 2017 all rights reserved

Scale selection – Quantitative studies

Rel

ativ

e Im

porta

nce

• Use established scales where possible

• Look for most recent versions of scales and most cited from top journals

• Combining multiple scales is NOT using existing measures (you are creating a new scale)

• Take care not to alter scale too much (slight changes OK)

• Take care with changes to scale measures (e.g. 5-point or 7-point items)

©Mazzarol 2017 all rights reserved

Validity in Qualitative Research

Source: Maxwell (1992)

• Descriptive validity– Factual accuracy of findings– Did people say what was said?

• Interpretive validity– The meaning of the findings

• Theoretical validity– Do the findings support theory?

• Validity of concepts used• Validity of relationships

• Generalizability– Can the findings be used again?

• Evaluative validity– Judgements of right & wrong

©Mazzarol 2017 all rights reserved

Group Exercise

• Working in groups go online and find reliable measurement scales for:– “Learning Orientation”– “Customer Satisfaction”

• Select the best measures for these constructs

• Justify your selections

©Mazzarol 2017 all rights reserved

Sampling

Sample

True populationA sample

representsthe true

population

©Mazzarol 2017 all rights reserved

Sample Size

• The more precise an estimate needs to be the larger the sample should be

• Statistical analysis requires a minimum of 30 respondents for reliability

• In populations where there is a high degree of variance in responses, this minimum number needs to be increased

• Studies that wish to examine the individual responses of sub-groups within a sample need to be larger than 30

• Samples must be representative of the true population they represent

©Mazzarol 2017 all rights reserved

How many respondents do you need?

Sources: Corbett (1991)

Desired accuracy level Desired confidence level

.01 .05

1% 16,587 9,604

2% 4,147 2,401

3% 1,843 1,067

4% 1,037 600

5% 663 384

Note: The table assumes maximum variability for a binomial variable

©Mazzarol 2017 all rights reserved

Video – 10 Step Guide to Sampling

©Mazzarol 2017 all rights reserved

10 Step Guide to Sampling

Sources: RCU (2010)

1. What do you want to know?

2. Who is the population?

3. Census or sample?

4. Generalising from a sample – keep it representative

5. Choosing a sample method – simple, stratified, quota

6. The sampling frame – who are the target population?

7. Sample size – depends on accuracy, aims and size of target pop.

8. Sampling error – difference between sample & true pop.

9. Correcting bias – statistical techniques “weighting” can help correct

10. Sampling plan – target population, size, method, sampling frame

©Mazzarol 2017 all rights reserved

Sampling plan design

• Probability Samples– Simple random sample (total

population)– Cluster sample (segments of

total population)– Stratified sample (segments with

common features)

• Non-Probability Samples– Quota samples (selection on

common criteria)– Convenience sample (selection

on availability)

©Mazzarol 2017 all rights reserved

Survey research design

• A very common method for business management and marketing studies.

• Key principles:– Establish the aims of the research.– Identify the target population who will be studied.– Identify the best way to collect reliable data:

• Online• Telephone• Mail • Face to face

– Decide how large a sample should be– Determine the most appropriate questions to ask:

• Open ended• Multiple choice • Type of scales to be employed

– Choose the type of analysis BEFORE you collect the data.

Source: Shuttleworth (2008)

©Mazzarol 2017 all rights reserved

Principles of questionnaire design

• Use simple every day language• Keep questions short and avoid double-barreled

questions– “Do you think women and children should be given the

first available flu shots ?”• Avoid double negatives

– Please tell me whether you agree or disagree with the following statement about teachers in the public schools:

• Teachers should not be required to supervise students in the halls, the lunchroom, and the school parking lot

• Lay out the questionnaire in a way that is easy to read and leads the respondent through the questions

©Mazzarol 2017 all rights reserved

Video – 10 Step Guide to Questionnaire Design

©Mazzarol 2017 all rights reserved

10 Step Guide to Questionnaire Design

Source: RCU (2010)

1. Keep it simple and make it interesting - Short, clearly worded items, important items up front

2. Keep it short - Don’t add unnecessary items3. Use the language the respondents would use - Pre-test (pilot) with

sample of target audience4. Think about the order of the questions - How one item is responded to

can influence another that follows5. Avoid subjective terms - Don’t use terms that are open to interpretation6. Don’t worry about including a middle response option if it reflects what

some respondents would want to say - Middle option scales are OK7. Avoid double negatives - Keep lead-in stems positive rather than negative

with “agree-disagree” items8. Make scales logical - Ensure that any scale numbering “makes sense” to

the respondent (e.g. 10-point scales versus 5-point scales)9. Don’t forget zero! - If you use 10-point rating scale include a zero option

(e.g. 0-10 option)10. Put the personal questions at the end - Personal items (e.g. age,

ethnicity) can put people off responding or bias

©Mazzarol 2017 all rights reserved

Writing good questions

• Ensure that the respondent:– Understands what is being asked– Can meaningfully comment on the subject

being asked• Ensure that the question records findings in the

most appropriate way:– Likert-type rating scales

• Strongly disagree 1 2 3 4 5 Strongly agree

– Dichotomous:• Yes/No or Agree/Disagree

– Open ended:• What do you think about….

©Mazzarol 2017 all rights reserved

Design and Layout

• Ensure that the respondent understands what the questionnaire is designed to achieve

• Structure the questionnaire logically• Use white space and consistent type face• Ensure that questions flow logically and in

order that won’t conflict– General questions should precede specific

ones• Allow sufficient room for answers• Make sure instructions are clear

– “If NO, go to Question X”

©Mazzarol 2017 all rights reserved

Methods of Data Collection

• Personal Interviews:– Provide a powerful source of information allowing in-depth questioning but are

expensive and require skilled interviewers• Telephone Interviews:

– A faster method of reaching a target group, response rates are generally high, but limited by cost to shorter questionnaires with simple wording.

– Requires special training of data collection– Ensure that CATI is used if possible– Telephone and fax combinations are a useful alternative

• Mail Surveys:– Self-administered questionnaires enable more lengthy and complex questions– Lack of control and slow turnaround– Response rates usually 10%

• Online Surveys:– Fast & getting more reliable with large, national panels– Lower cost compared to telephone surveys– Allows use of multimedia display if required– Automatically codes

©Mazzarol 2017 all rights reserved

Boosting response rates

• Saliency – how relevant is the survey to the respondent?

• Pre-notification – letting the respondents know that they will be surveyed

• Post-notification – follow up after a survey has been distributed

• Payments or incentives • Use of pre-paid envelopes & endorsements• Use of good covering letter• Overall “get up” of the questionnaire• Telephone survey scripts

©Mazzarol 2017 all rights reserved

How to fail your PhD/DBA

• Don’t talk to your supervisor about examination:– Start to look for examiners as early as possible.– Consider what they will be expecting and why.

• Have the thesis examined by in-expert people:– Inexperienced or inappropriately qualified examiners can

be lethal. • Write your introduction first

– Examiners read the abstract, introduction and conclusions first, then look at the references.

– Questions raised in the introduction should be answered in the conclusions.

• Write a bad literature review– Like a “party frock” for the thesis the “little black dress”.– Cover the major authors with a coherent thread that

connects with the purpose of the study.• Don’t let anyone else do your copy-editing

– Don’t include typos, grammatical or spelling errors, poor formatting or sloppy footnoting or referencing.

Source: Mewburn (2013)

©Mazzarol 2017 all rights reserved

Writing the proposal

T• Topic sentence controlling idea

• One key idea for every paragraph

E• Explain, elaborate, or define

• Use plain English and short sentences• Back up your thesis and save frequently

E• Evidence, examples or illustrations

• Support all assumptions or conclusions with evidence

Source: Massey University

©Mazzarol 2017 all rights reserved

Developing the thesis – step 1

• Getting organised:– Commence by writing the sections you

know best– Rewrite the proposal into chapters for

the thesis – Use real names & places in early drafts

for cases– Clearly identify each version of the

thesis– Develop diagrams & models by hand

at first• Make use of headings & tables or

figures but:– explain these in the text– be consistent with descriptions

Source: Levine, 2008

©Mazzarol 2017 all rights reserved

Developing the thesis – step 2

• Review other theses prior to commencement• Write clearly & avoid ambiguity & use plain

English• Use the table of contents to help improve the

manuscript• Make conclusions “real” don’t just summarise

findings• Make future implications for researchers

meaningful• Finish Chapter 1 last

Source: Levine, 2008

©Mazzarol 2017 all rights reserved

Working with your supervisor

• Develop a “blueprint” for your thesis• Start writing initial chapters early• Make sure you keep to plan• Take notes during meetings & keep your

promises• Meet/communicate regularly• Remember that it is your thesis & your

responsibility

©Mazzarol 2017 all rights reserved

Planning the macro-structure of the thesis

• Key issues:– Make the structure accessible to readers – well organised,

logical and cumulative.– Keep your argument “on track” with each chapter linking to

the next.– You don’t need to write each chapter in sequence.– Have a clear plan of the thesis structure before you start.

• How long should it be?– Somewhere between 50,000 and 100,000 words.– That is around 200 to 330 pages of A4 paper double-spaced.– Don’t “over write” the thesis and make it too long.

• What should its structure be?– Each chapter should be around 6,000 to 7,000 words. – Keep each chapter approximately the same length.– Chapters should be logically structured.

Source: Dunleavey (2003)

©Mazzarol 2017 all rights reserved

Planning the macro-structure of the thesis

Source: Dunleavey (2003)

• Chapter 1: Introduction– Topic that this thesis is about; the research problem; why is this

problem important?; research questions; context of study; outline of remaining chapters.

• Chapter 2: Literature Review– Theoretical and conceptual foundations of the study; what is already

known; hypotheses emerging from this analysis.

• Chapter 3: Methodology– Research design; sampling; qualitative or quantitative.

• Chapter 4: Data Analysis and Findings– Data collected; testing and analysis; summary of findings.

• Chapter 5: Discussion of Findings– Explanation of the findings; linking back to literature.

• Chapter 6: Conclusions, Implications, Limitations– What does it mean for theory, practice, policy; limitations of your

research; future research directions.

©Mazzarol 2017 all rights reserved

Structuring a chapter

Source: Dunleavey (2003)

Introductory text200 to 1,000 words [no sub-heading]

3.1 First main section2,000 to 2,500 words [first order heading]

3.2 Second main section2,000 to 2,500 words [first order heading]

3.3 Third main section2,000 to 2,500 words [first order heading]

3.4 Fourth main section2,000 to 2,500 words [first order heading]

Conclusions200 to 1,000 words [second order heading]

Avoid using two many levels of sub-heading – no more than three.

Introduce each chapter with a short summary as to what it is about.

Keep sections about the same length and organise around key themes.

If you use lengthy quotes, place them into a separate paragraph and indent them –perhaps use a separate font size or font.

Keep consistent with formatting, headings and the labelling of graphs and tables.

Use a chapter numbering system.

Make it easy to read and articulate through the thesis.

Conclude each chapter with a summary of what was covered and where you will go with the subsequent chapter.

©Mazzarol 2017 all rights reserved

Ethics Approvals

Review the UWA Ethics Guidelines & Materials

– What are the key things you must do?– Will you be likely to encounter any ethical

issues?

©Mazzarol 2017 all rights reserved

Group Exercise

• Working in groups go online and find the UWA Human Ethics Office:– “Obtaining ethics approval”– “How to write a good ethics

application”– HREO application form

• Make a list of the key things you will need to consider for your application.

• Discuss common areas of concern within your group.

©Mazzarol 2017 all rights reserved

Video – Preparing your thesis defence

©Mazzarol 2017 all rights reserved

Next steps

• Work closely with your supervisors

• Draft due 16 June 2017

• Final submission due 14 July 2017

• DBA thesis proposal defence 29-30 July 2017

©Mazzarol 2017 all rights reserved

Thinking Exercise – Prepare a “To Do” list

• Review the key things you must do now to get your project proposal ready

• Make a “To Do” list• Discuss with your group and report

back to the class

©Mazzarol 2017 all rights reserved

End of presentation