Week 1 Exam Flashcards
Why talk about research? Why research is important:
Managers need to make decisions all the time, some are based on intuition, experience, or advice. However, when there is unexplored knowledge needed, research might be the only option
Understanding the principles of good research might help to:
1) To execute (or commission) own research
2) To critically understand research done by others
Basic Steps in Research:
1) Formulate knowledge question
2) Collect relevant knowledge that is out there
3) Collect new, additional data
4) Analyze and interpret
5) Formulate the answer to the question
Main ingredients of science (recipe):
1) Theory - formalized explanation of a phenomena
2) Expectations - propositions of hypotheses what we expect to observe
3) Studies - Experiments, surveys, case studies, secondary data analysis, simulations
4) Observations - data stemming from studies (that need to be analyzed and understood)
Three logics:
1) Inductive - given a series or observations, a more general explanation or theory is derived.
Logic - observations are used to generate logical but not yet tested conclusions
Generalizability - from specific (observations) to general
Data is used to explore phenomena, identify patterns and create theory.
Theory-building, exploratory
2) Deductive - Based on premises that are true we derive logical conclusions
Logic - when observed premises are true, the conclusions must be true.
Generalizability - from general theories to specific
Data is used to evaluate propositions or hypotheses from existing theories
Theory testing
3) Abductive - Based on interactions between observations and theories, we come to a likely explanation for what we
see. Often used after a surprising phenomena is observed, then theory is examined, and conclusions made:
I know two letters (observation), based on these two letters the word could be apple (theory that only apple starts with ap).
Logic - Observations are used to generate assumptions which are tested
Generalizability - interaction between specific and general
Data is used to locate the phenomenon in the theory, by identifying themes and patterns, then this theory is modified or extended
Theory building, modification
Logics can be combined in a single study:
* Abduction to conceive ideas
* Deduction to logically construct propositions or hypotheses
* Induction to observe reality and generalize
However, there is usually overarching logic.
Research cycle:
States that research process is circular rather than linear:
1) Management problem
2) Knowledge question
3) Review of evidence
4) Research design
5) Data collection
6) Data analysis
7) Research outcomes
8) Recommendations to management
Throughout each step critical reflection is applied. (see next flashcards for deep-dive in each step)
1) Management problem (research cycle):
From a problem “Mess” the most critical and relevant issue is identified. The issue is related to practice and performance
Different stakeholders and their perceptions on the issue are considered, what impacts it and what is the overarching issue. Also allows to recognize possible sponsors of research
Define initial knowledge questions behind the issue
Through abductive reasoning gather relevant observations and conduct quick exploratory research
Define research objective
2) Knowledge question (research cycle):
Start collecting research and evidence around the initial set of knowledge questions
Develop and refine the knowledge questions, so that there is one main question and 3-5 sub questions
In case of theory testing also hypotheses are developed
Types of knowledge questions:
Descriptive - what is happening, explores what are the benefits of something, results, and situation, mainly related to exploratory research type
Explanatory - why something is happening the way it is, why things are the way they are, mainly related to theory-building research
Predictive - how things will happen, mainly related to theory-testing research
Prescriptive - how things should be done, mainly related to decision science
Also in medicine, evaluative knowledge is possible.
3) Review of evidence (research cycle):
Some evidence and literature have already been reviewed in the previous step
A systematic search for literature is initiated based on questions developed:
Academic and professional literature is reviewed
A literature review is developed
4) Research design (research cycle):
Determine the type of research based on knowledge question: exploratory, theory-building, theory-testing, decision science
Determine what strategies used:
Quantitative - surveys, experiments, secondary data analysis (mainly used in theory-testing)
Qualitative - cases, interviews, focus groups (in theory - building and exploratory research)
Consistency between research type, question, objective is established.
Research design should establish
1) Data collection
2) Data analysis
3) Threats to validity and how to resolve this
4) Timing of the project
7) Research outcomes (research cycle):
Results section: represents numeric findings and overall results of the study. Makes no interpretation of the results
Discussion: interprets the results and puts it into words, consists of:
1) Conclusion - answers the research questions and if the research objective is achieved
2) Contribution to the theory - how the research adds to the current state of knowledge
3) Contribution to the practice - mainly related to decision science
4) Limitations - an acknowledgment of limitations that impact the research outcomes and the weaker parts of research.
5) Suggestions for future research
8) Recommendations to management:
To support decision-making in the organization.
Can the recommendations be captured in a CIMO statement: In this context, what interventions can be made, so that mechanisms, lead to certain outcome
Must be careful to not over-generalize, and clearly identify advantages and disadvantages, risks, and critical success factors related to the recommendation.
Critical reflection (research cycle):
To be critical it does not mean to destructively criticize, but rather to think critically:
1) Identify strong parts and weak parts of the research
2) Search for agreements and disagreements (contrasts) in the research and current state of knowledge.
3) Be mindful of biases and quality of sources used
4) Reflect on whose voice is “amplified” (research) and “muted” (not recognized as often in the knowledge)
5) Reflect on the ethics of your research and used articles
Data vs Knowledge:
Data - refers to raw observations and information that is unprocessed
Knowledge - data that is transferred into knowledge through “processing” - reflecting, analysis, interpretations, linked to other pieces of knowledge. Knowledge is meaningful, data does not have to be.
Primary data vs Secondary data:
Primary data: Obtained by the researcher.
1) Always raw and needs processing
2) usually is more expensive to obtain and takes more time
3) Smaller sample
4) Has higher quality as is more closely related to the research
5) Collected through questionnaires, surveys, interviews, photos, videos and experiments
Secondary data: taken from sources that have already prepared the data.
1) Already processed and ready for use
2) Often cheap and fast access
3) Larger samples
4) Not as related to research and is taken from different contexts
5) Collected from websites, industry statistics, reports, existing databases
Often, difficult to assess what is primary and secondary.
Rule of thumb: is the measurement approach (how am I going to analyze the data) developed by you or not?
Quantitative vs Qualitative data:
Qualitative data:
Non-numeric data that is gathered in an unstructured format.
Usually more subjective depends on the interpretation
The goal is to categorize or create themes
Includes: words, videos/images
Quantitative data - numeric values, that are continuous or discrete
Questions include, how many, often, etc.
Objective data, that is not dependent on interpretation.
Statistical analysis is used
Include secondary data, surveys, measurements, and questionnaires.
Simulated vs Empirical data:
Simulated data: Simulation is the creation of a model that can be manipulated logically to decide how the ‘real’ physical world works
1) It can be used to see how the developed model works or explore alternatives (often done at the beginning of the research)
2) Faster and cheaper to obtain
3) Used for predicting future events, as empirical data looks at past or present
4) Based on meaningful and real data
Results of empirical and simulated might not be the same:
1) Error in the measurements
2) Errors in modeling of simulation
3) errors in the formulation of hypotheses
Exploratory research:
The outcome is a description of a phenomenon (categorization, text of explanation, a process map). Leads to descriptive knowledge.
1) Management problem - the problem not defined well yet
2) Knowledge question - What is the impact of procurement digitization on the operational procurement processes?
What are the common benefits and risks of procurement digitization?
3) Review of evidence - practically no literature in the field, can be drawn some conclusions from relevant works in other fields, exploration of practice (professional literature)
4) Research design - usually qualitative but can be quantitative as well. Strategies include - interviews, focus groups, case studies, surveys. Logic is inductive or abductive (if there is some theory).
7) Research outcomes - explanation of a phenomenon
8) Recommendations - no strong recommendations can be made, however, helps to understand the problem.
Theory - building:
Output are theoretical proposition that
explains certain phenomena or processes. Leads to explanatory knowledge
1) Management problem - the problem is new but clear: E.g., procurement digitization has been adopted by many sectors and companies, but it is not clear whether it also can be useful for supplier relationship management
2) Knowledge question - How can supplier relationship management be enhanced with introduction of digital technologies?
3) review of evidence - there is some literature available, but is very limited, with no theory involved. Useful to identify variables, concepts. Business literature used.
4) Research design - likely to be inductive or abductive. Strategies include - interviews, case studies, focus groups, secondary data (multi-study methods when combined)
7) Outcomes - Assumptions (theoretical propositions) about how procurement digitization is related to various practices in SRM; generalizability is always a potential issue
8) Theory-building research cannot provide strong advice (propositions have not been tested), but may be used as an attempt to explain how something works; recommendations can be very contextual
Theory - testing:
The output is ‘proof’ and quantifications of
relationships between the established
variables, generates predictive knowledge.
1) Management problem - here is theoretical knowledge (assumptions) about the topic, but not yet verified/proven.
2) Knowledge question - hypotheses developed
3) Review of evidence - available considerable amount of academic literature that is related to the topic and is relevant. There is no relationship established that is specified in the research
4) research design - deductive, starts with theory. Strategies are quantitative - surveys, secondary data, questionnaires, and experiments.
7) Research outcomes - confirmation or rejection of the hypothesis
8) Recommendations - strong recommendations can be made, however, must be aware of context, but perceived as generalizable
Decision science:
The output is techniques, algorithms for
decision-making, optimization models. Leads to prescriptive knowledge.
1) Management problem - The topic can be already studied in general but represents a particular management problem for the company.
2) Knowledge question - forecasts, mathematical models, how to improve processes
3) Review of evidence - business literature and academic articles
4) Research design - uses inductive or abductive logic. Strategies include mathematical modelling, and simulation.
7) Outcomes - Mathematical tools for modelling and forecasting
8) Decision-support tools with recommendations on how to use them are provided. They are likely to be highly contextual (if built with the company information) and thus relevant for the business; for use in other contexts the models may have to be changed.
