Lecture note on quantitative research methods design quality: for MBA students
Highlight 3 main quantitative research methods design
criteria and offer 3 practice advices on how to improve the quantitative
research methods design in the context of doing a part-time MBA dissertation
project.
For a part-time MBA dissertation, the three main quantitative
research design quality criteria are:
|
Criterion |
What
it means |
|
Internal
validity |
Whether your results truly
come from the variables you're testing, not from confounding factors or
design flaws |
|
External
validity |
Whether findings can be
generalized beyond your sample to a wider population or other settings |
|
Reliability |
Whether repeating the same
study with the same measures would produce similar results |
These criteria address whether your findings are true, generalizable,
and consistent.
3 Practical Advices to Improve
Quantitative Design for a Part-Time MBA Dissertation
1. Use validated measurement tools and pre-test your questionnaire
Where possible, adopt established scales from peer-reviewed literature rather
than creating new questions. Pre-test structured questions with a small group
to ensure respondents understand the wording and there are no comprehension
problems. For a part-time student with limited time, this reduces the risk of
costly data collection errors later.
2. Ensure appropriate sample size and use probability (or targeted
non-probability) sampling
Your sample must be large enough for statistical analysis. Clearly define your
target population and sampling method upfront. For part-time MBA students who
may struggle to access large random samples, use targeted sampling of
relevant professionals in your industry while acknowledging this limits
external validity. Document your sampling rationale to justify reliability and
validity.
3. Plan for confounding variables and choose statistical tests that
match your research questions
Identify potential confounding variables before data collection and controls
for them in your design. Understand why a particular test
(e.g., t-test vs. chi-square) fits your data and questions—not just memorizing
procedures. Align research questions directly with objectives and your
theoretical framework. For part-time students balancing work, this upfront
planning prevents costly re-analysis later.
Part-time student tip: Given time constraints, consider data availability early and keep
your questionnaire intentionally short, focusing only on questions that change
decisions. Justify your quantitative choice clearly to supervisors—explain why
it fits your research question, data, and objectives.
Please clarify your ideas a bit with regard to the
part-time MBA dissertation research theme of "to evaluate how corporate
innovation capability, corporate competitive advantage and corporate strategic
aggressiveness are related in the Hong Kong Insurance sector via the lens of
positivism and using the research methods of questionnaire survey and
structured interviews with 5 senior managers in the Hong Kong insurance
sector".
Your research
theme is quantitative-focused (positivist lens, questionnaire
survey) but also includes a small qualitative component (structured
interviews with 5 senior managers). Here's how to clarify and strengthen the
design:
1. Clarify the Research Philosophy Alignment
Positivism means:
·
You assume there
are objective, measurable relationships between innovation
capability, competitive advantage, and strategic aggressiveness
·
You seek generalizable
laws about how these variables interact in the Hong Kong insurance
sector
·
You maintain researcher
distance from subjects and use numerical, statistical analysis
The tension to
address:
Structured interviews with 5 managers are typically interpretivist (qualitative,
exploratory). Under positivism, you must justify them as auxiliary
quantitative support, not primary qualitative inquiry:
·
Use interviews
to validate questionnaire items before survey deployment
(pre-testing)
·
Use interviews
to confirm numerical patterns found in survey data
(triangulation)
·
Keep interview
questions structured and closed-ended (not open exploratory)
2. Clarify Your Primary vs. Secondary Methods
Recommendation:
Make questionnaire survey your PRIMARY method
|
Method |
Role |
Sample
Size |
Justification |
|
Questionnaire survey |
Primary data collection |
100–200 insurance
professionals |
Provides statistical power
for testing relationships |
|
Structured interviews |
Secondary/validation |
5 senior managers |
Pre-tests questionnaire
items; triangulates survey findings |
Why this matters
for part-time MBA:
·
5 interviews alone
cannot establish statistical relationships between variables
·
Questionnaires
give you internal validity (true causal relationships)
and reliability (repeatable results)
·
Interviews add
credibility but should not drive your main conclusions
3. Clarify Your Variable Definitions &
Measurement
You need to
operationalize three variables into measurable constructs:
|
Variable |
How
to measure (example scales) |
Source |
|
Corporate
innovation capability |
Likert-scale items on R&D
investment, new product development, innovation processes (e.g., 5-point
scale: "Our company regularly develops new insurance products") |
Use validated scales from
peer-reviewed innovation literature |
|
Corporate
competitive advantage |
Measures of market position,
profitability relative to competitors, customer retention rates, brand
strength |
Adopt established
competitive advantage scales |
|
Corporate
strategic aggressiveness |
Items on market entry speed,
pricing aggressiveness, competitive response intensity, risk-taking in
strategic decisions |
Use validated strategic
aggressiveness scales |
Critical step: Pre-test these questionnaire items with
the 5 senior managers to ensure clarity and reliability before full survey
deployment.
4. Clarify Your Sampling Strategy for Hong
Kong Insurance Sector
Target population: Insurance companies operating in Hong
Kong (life, non-life, reinsurance)
Sampling approach
for part-time student:
·
Probability
sampling is ideal for
external validity but often impractical
·
Use targeted/non-probability
sampling: contact insurance professionals through industry networks,
professional associations, or your workplace
·
Document
limitations: Acknowledge that
non-random sampling limits generalizability but is justified for MBA dissertation
scope
Sample size
guidance:
·
For
correlation/regression analysis (testing relationships between 3 variables):
minimum 100 respondents for statistical power
·
For more complex
analysis (e.g., multiple regression): aim for 150–200
5. Clarify Your Statistical Analysis Plan
Since you're
testing how three variables are related, your analysis should
include:
|
Analysis
Type |
Purpose |
When
to Use |
|
Pearson
correlation |
Test strength/direction of
relationships between each pair of variables |
Initial exploration |
|
Multiple
regression |
Test whether innovation
capability AND strategic aggressiveness predict competitive advantage |
Main hypothesis
testing |
|
Mediation
analysis (optional) |
Test if strategic
aggressiveness mediates the innovation→competitive advantage relationship |
If your theory suggests this
pathway |
Positivist
requirement: Your
statistical tests must directly answer your research questions. Justify each
test choice theoretically, not just procedurally.
6. Clarify Internal Validity Controls
To ensure your
findings reflect true relationships (not confounding factors):
|
Potential
Confounder |
Control
Strategy |
|
Company size (larger firms
may have more innovation resources) |
Include company size as a
control variable in regression |
|
Firm age (older firms may
have different competitive positions) |
Include firm age as control
variable |
|
Insurance sub-sector (life
vs. non-life may differ) |
Include sub-sector as
control or stratify analysis |
|
Respondent role (senior vs.
middle management perspectives) |
Ensure sample includes mix
of roles; control for role in analysis |
Part-time student
tip: Keep your
questionnaire short—include only questions that directly test your hypotheses
or control for confounders.
7. Revised Research Statement (More Precise)
Original: "to evaluate how corporate
innovation capability, corporate competitive advantage and corporate strategic
aggressiveness are related in the Hong Kong Insurance sector via the lens of
positivism and using the research methods of questionnaire survey and structured
interviews with 5 senior managers in the Hong Kong insurance sector"
Sharpened version:
"To
quantitatively examine the relationships between corporate innovation
capability, competitive advantage, and strategic aggressiveness in the Hong
Kong insurance sector, using a positivist research philosophy. Primary data
will be collected via a structured questionnaire survey (n≈150 insurance
professionals), with structured interviews of 5 senior managers serving as a
pre-test for questionnaire validation and triangulation of survey
findings."
8. Key Supervisors Will Question (Prepare
Answers)
|
Question |
Your
Response |
|
Why positivism, not
interpretivism? |
You seek objective,
generalizable relationships between measurable variables, not subjective
interpretations |
|
Why only 5 interviews if
qualitative? |
Interviews are structured,
closed-ended, and serve survey pre-testing/triangulation—not primary
qualitative inquiry |
|
How will you ensure internal
validity? |
Control for confounders
(company size, age, sub-sector) in regression analysis |
|
How will you ensure external
validity? |
Acknowledge sampling
limitations; frame findings as applicable to Hong Kong insurance sector with
caution about generalizability |
|
How will you ensure
reliability? |
Use validated scales from
literature; pre-test questionnaire; document procedures for replication |
Final Recommendation for Part-Time MBA
Context
Given your time
constraints:
1.
Focus 80% effort
on questionnaire survey (primary
quantitative method)
2.
Use 5 interviews
strategically: pre-test
questionnaire items + 1–2 follow-up interviews to triangulate key survey
findings
3.
Keep statistical
analysis straightforward: correlation +
multiple regression (not overly complex mediation/moderation unless
theoretically essential)
4.
Justify choices
clearly to
supervisors: explain why quantitative fits your research question, data, and
objectives
This design
maintains positivist rigor while being feasible for a part-time student
balancing work and dissertation.
** A collection of lecture notes on the subject of research methods for MBA students, 2026 June
A collection of blog notes on using chatgpt for research purpose.
No comments:
Post a Comment