ARM lecture 1

Why do we talk about research?

Managers need to take decisions all the time. If new knowledge is required to make a decision, research may be needed. Understanding the principles of good research helps.

5 basic steps in research

1. Formulate a knowledge question

2. collect relevant knowledge that’s already out there

3. collect new, additional data

4. analyze and interpret

5. formulate the answer to the question

Main ingredients of science

• Theory

• Expectations

• Studies

• Observations

Inductive reasoning

A logical process where specific observations lead to general conclusions (theory building), typically not tested. For example, if all observed swans are white, it concludes that all swans are white. (identify themes/patterns/phenomenon to create a theory)

Deductive reasoning

A logical process that starts with a general statement or hypothesis (all the observed premises) and examines the possibilities to reach a specific, logical conclusion. This approach is used to test theories by applying them to specific cases.

Abductive reasoning

Observations generate assumptions, which are tested (logic). Generalizability connects specific cases to broader concepts. Data explores phenomena, identifies patterns, and extends existing theories. This process aids in theory building or modification (elaboration).

Research cycle

• Managerial problem

• Knowledge question

• Review of evidence

• Research design

• Data Collection

• Data analysis

• Research outcomes

• Recommendation to management

It’s a cycle with in the middle, critcial reflection.

Managerial problem

• Typically a performance problem

• often part of a mess, multiple interrelated problems

• define the knowledge questions behind the managerial problem

• use abductive reasoning, collect prior studies around the problems and do some quick exploratory research

• bring everything together in a research objective

Knowledge question

• start collecting literature around the concepts

• refine the questions until you have a main research questions and subquestions

• if theory-testing, hypotheses

• four types of research questions

  • Descirptive knowledge(how things are)

  • Explanatory knowledge(why things are that way)

  • Predictive knowledge (how things will be)

  • Prescriptive knowledge (how things should be done)

Review of evidence

• you started collecting literature

• start more systematic search for relevant literature (Academic and professional literature)

• capture results in critical literature review

Research design

• determine research type

• check the research objective, questions and type (are they consistent with each ohter)

• determine research strategy

• research design describes: plan for data collection, analysis, threaths to validity and how to deal with those, time plan/project plan

Research outcomes

• results: outcomes of your analysis

• discussion: conclusion, contribution to theory, contribution to practice, limitations, suggestions for future research.

recommendation to management

• CIMO statement (Context, Intervention, Mechanism, Outcome)

• be careful not to over-generalize, be mindful of context

• think about advantages, disadvantages, risks and critical succes factors related to a recommendation

Critical reflection

• critical is not to criticize everything you read, see or have produced yourself

• force yourself to identify stronger and weaker parts

• search for agreements and disagreements

• check the quality and possible bias of the literature/data source

• reflect on whose voice is amplified and whose voice is muted

• reflect on the ethics

Data and knowledge

Data is unprocessed inforamtion, raw form. Knowlegde is processed data, and always meaningful.

Primary and secondary data: when in doubt

Is the measurement approach (how am i going to analyze the data) developed by you or not?

Simulated and empirical data

Simulated data is generated by creating a logical model to explore or test proposed solutions, useful in early research phases for predicting outcomes. It is faster and cheaper to obtain than empirical data, which is based on real-time or historical events but cannot predict the future. While simulated data mimics the characteristics of real data, results from simulations and empirical studies may differ due to potential errors in measurement or modeling.

Exploratory research

• Management problem: new or not yet defined

• Knowledge question: impact/benefits etc

• Review of evidence: academic literature typically scarce, maybe borrowing literature from other topics

• Research desing: inductive logic (or abductive if not completely new). Strategies may be case study, secondary data analysis, interviews, focus groups, survey, combination of these

• Data collection : likely qualitative, quantitative also possible

• Data analysis: depending on the data type, both qualitative and quantitative approaches

• research outcomes: improved description

• Recommendations: exploratory studies can not give strong advice, but can help understand the problem and develop next steps. —> descriptive knowledge

Theory buidling research

• Management problem: new but clear problem, understanding main components and relationships is needed

• Knowledge question: How can they enhance etc.

• Review of evidence: Academic is limited, but maybe understand important concepts (variables), useful theories from other topics. Business may be avaiable and helpful

• REsearch design: inductive logci (or abductive for theory modification or elaboration)

• Data collection: Qualitative data is involved likely, quan can be

• Data anlayis: both qual and quan

• Research outcomes: Assumptions about how

• Recommendations: Theory building research not strong advice, not tested, but explain how something works. —> exploratory knowledge

Theory testing research

• Management problem: There is theoretical knowledge (assumptions), but not proven/verified.

• Knowledge question: Hypothesis

• Review of evidence: Academic literature is there, theory is fomrulated and maybe tested, no proof of relationship that form the topic

• Research design: Deductive logic . Possible strategies are surveys, experiments, maybe secondary analysis.

• data collection: quan

• Data analysis: quan

• Research outconmes: proof or rejection of effect

• Recommendation: likely to be convincing, evidence for manageral decisions. —> generates predictive knowledge

Decision science

• Management problem: already be studied but represent particular managment problem for the company

• forecast for example

• literature on models etc

• inductive or abductive logic. possible strategies, mathematical modelling

• data collection: quan

• data analysis: train test improve models

• research outcomes: mathematical tools

• Recommnedations: decision support tools with recommendation how to use them, likely hihgly contextual. —> prescriptive knowledge.