Week 1 Flashcards
Basic steps in research
- Formulate a knowledge question
ex. What product features are valued the most? - Collect relevant knowledge that’s already out there
ex. Knowledge on how to study this question (survey?)
ex. Existing lists of product features - Collect new, additional data
ex. Develop a strategy to sample respondents
ex. Respondents rate product designs in a survey - Analyze and interpret
ex. Perform relevant type of analysis
ex. Measure the utility of different product features - Formulate the answer to the question
ex. Rank product features and product designs based on relative utility.
Main ingredients (or recipe) of science
- Theory: formalized explanations of phenomena
- Expectations: propositions or hypotheses about what we expect to observe
- Studies: Experiments, surveys, case studies, secondary data analysis, simulations
- Observations: Data stemming for studies (that need to be analyzed and interpreted).
Scientific reasoning; three main logics
- Inductive reasoning
* Given a series of observations, we derive an explanation / generalization that is probably true
*From specific to general
ex. All observed swans are white;
therefore, all swans are white - Deductive reasoning
* Based on premises that are true, we logically come to a conclusion that is true
*From general to specific
ex. All men are mortal. Socrates is a
man; therefore, Socrates is mortal. - Abductive reasoning
* Based on interactions between observations and theories, we come to a likely explanation for what we
see.
*From the interactions between specific and general
ex. Wordle game
Combination of logics- possible?
Yes.
Logics are often combined in scientific argumentation:
* Abduction to conceive ideas
* Deduction to logically construct propositions or hypotheses
* Induction to observe reality and generalize
- For example:
Induction to identify patterns, followed by abduction to develop a solution for a problem
Nevertheless, a study as a whole often has one overarching logic
Describe research cycle
<Critical>
Managerial problem--->
Knowledge question--->
Review of evidence--->
Research design--->
Data collection--->
Data analysis --->
Research outcomes --->
Recommendation to management
</Critical>
Research cycle: Managerial problem
Key Aspects:
Typically a performance issue (which dimension is unsatisfactory?)
Often part of a problem mess with multiple interrelated problems and complex cause-effect relationships
Steps to Address:
Define knowledge question(s) behind the problem
Use abductive reasoning and review prior studies
Conduct quick exploratory research
Formulate a research objective – What do you aim to achieve?
Tip: Clarifying the root issue leads to a more effective research direction.
Research cycle: Knowledge question
Gather literature based on research objective & initial questions
Refine into 1 main research question + 3-5 sub-questions
For theory-testing research, formulate hypotheses
- Four types of research questions as related to the type of knowledge they generate (increasing levels of
complexity):
1. Descriptive knowledge (how things are)
2. Explanatory knowledge (why things are that way)
3. Predictive knowledge (how things will be)
4. Prescriptive knowledge (how things should be done)
Research cycle: Review of evidence
Key Aspects:
*Expand literature search based on refined research questions
*Conduct a systematic search for relevant sources
Types of Literature:
1.Academic – Journal articles, scientific books, working papers, conference papers
2.Professional – Magazine articles, professional books, blogs, websites, newspaper articles
Final Output:
Critical synthesis of the literature, also known as a literature review
Research Cycle: Research Design
Key Steps:
*Determine research type – Exploratory (descriptive), theory-building, theory-testing, or decision science
*Ensure consistency between research objective, research questions, and research type
*Select research strategy – Experiment, survey, case study, modeling, etc.
Research Design Includes:
Data collection plan
Data analysis plan
Threats to validity & mitigation strategies
Time plan / project plan
Research Cycle: Research Outcomes
Key Distinction:
Results – Outcomes of your analyses
Discussion – Interpretation and implications of results
Five Key Discussion Elements:
1.Conclusion – Answers to research questions
2.Contribution to theory – New insights beyond existing literature
3.Contribution to practice – Practical applications (covered separately)
4.Limitations – Weaknesses of the research and their impact
5.Future research – Key follow-up questions
Research Cycle: Recommendation to Management
*Clearly link research findings to managerial decision-making
*Use a CIMO statement:
“In Context (C), if Intervention (I) is applied, Mechanism (M) will lead to Outcome (O).”
*Avoid over-generalization; consider the specific context
Evaluation Factors:
Advantages and disadvantages
Risks involved
Critical success factors
Data: Primary vs Secondary
–> is it developed by you or not
PRIMARY (“NEW”)
Collected by the researcher for the first time, with a particular research problem in mind
* Always raw and needs processing and analysis
* Often time-taking and expensive to collect
* Typically, smaller data samples
* Potentially higher quality for research (most of the
time all required variables are present)
SECONDARY (“OLD”)
Collected by any person or organization in the past for some other purpose
* Already somehow organized
* Often cheap and fast access
* Typically, bigger data samples and sets
* Potentially lower quality (often needs adjustment, not all variables or unnecessary variables present)
Data: Quantitative vs Qualitative
QUALITATIVE
* Words, pictures
* Descriptions of categorizations
* Questions What? How? Why?…
* Gathered by talking or observing =
subjective, open to interpretation
* Analyzed by reducing
QUANTITATIVE
* Numbers
* Counting or continuous measurements
* Questions How many (much)? How often?…
* Gathered by measuring and counting = universal, objective
* Analyzed statistically
Data: empirical vs simulated
Simulated data: Generated through models to mimic real-world scenarios, used for predictions and testing solutions.
Empirical data: Real-world, historical or real-time data, used to analyze past or current phenomena, but not for future predictions.
Name research types (research approaches)
1.Theory testing
2.Theory building
3. Exploratory/descriptive
4. Decision science
Theory testing research
The output is ‘proof’ and quantifications of relationships between the established variables.
LEADS TO PREDICTIVE KNOWLEDGE!!
–>There is theoretical knowledge (assumptions) about the topic, but not yet verified/proven.
–>Hypothesis can be formulated
–>Academic literature is present, but no proof on the research question
–>Deductive logic
–>Quant. data
–>Recommendations likely to be “convincing”
evidence for management decisions
Theory building research
Output are theoretical propositions that explain certain phenomenon or process
LEADS TO EXPLANATORY KNOWLEDGE!
–>problem is new but clear, understanding components and relationships.
–>Academic literature is limited, potentially related to other empirical contexts
–> Business literature is likely to be available and helpful
–>Inductive logic
–>Qualitative data is likely to be involved, but quantitative data also possible (for theory refinement)
–>Research outcome: assumptions
generalisability issue!!!
Cannot give strong advice but can attempt to explain how something works!
Exploratory research
Output is a description of a phenomenon or a process (maybe a text, a categorization, a process map, etc.)
LEADS TO DESCRIPTIVE KNOWLEDGE!
–>problem not yet well defined
–>Academic literature on new topics is typically scarce; some relevant theories from other topics can
be borrowed
–>Inductive logic
–>Qualitative data- likely, quant. data in for example survey.
–>Research outcome: improved description regarding an issue
cannot give strong advice! next steps of research
Decision science
The output is techniques, algorithms for decision-making, optimization models…
LEADS TO PRESCRITPIVE KNOWLEDGE!!!
–>Topic can be already studied in general but represent a particular management problem for the company.
–>Literature available
–>Inductive or abductive logic
–>Quant. data empirical + simulated data
–>Decision-support tools with recommendations on how to use them are provided.
–>Highly contextual dependant
