Research Methodology and Descriptive Statistics summary Units 1-22
Q3+4 Pre-Master Business Administration, SPSS not included in summary
Empirical research: systematically answering empirical questions
• Systematically: exclude the possibility that other answers are better than the answer we give
(follow a protocol).
• Empirical: about things we can observe
Procedure: relation between thinking and answers (e.g.: failed exam)
1. Thinking (=theory): Did not study hard enough.
2. Planning (=research design): Compare exams I did and see if studying more is working.
3. Observing (=data collection): Use my calendar and the grade centre to check that.
4. Analysing (=data analysis): Wow, yes! Only when I studied hard enough, I passed the
exam.
Wheel of Science: not a logical sequence of steps: anything goes. It does not matter at which step
you start. What does matter is that it is a way to defend conclusions: logic (a logical sequence of
steps). E.g.:
1. Question: ‘what causes cholera?’
2. Theory: caused by polluted water?
3. Research design: compare households
4. Collect data
5. Analysis data
6. Conclusion: it is indeed caused by polluted water.
Pagina 1 van 22
,Empirical research questions: are often asked in the context of decision making and design. They can
only be answered using observations.
‘How to’ questions are empirical, but can’t easily be answered. These ‘How to’ questions can be
broken up into multiple empirical research questions (ERQ).
E.g.: ‘How to reduce the number of people that smoke?’.
Types of questions: 3 types of questions: normative, conceptual and empirical. Different types op
research questions means different types of procedures -> need to distinguish ‘types of research
questions’.
1. Normative (what ‘should’ be the case?)
- Often starts with ‘Should we…?’, ‘Is it justifiable?’
- Not asking for legal facts
- Cannot be answered using observations only
- E.g.: ‘Should we introduce the death penalty in order to reduce crime?’
2. Conceptual (what does it mean?)
- Often starts with ‘what is’
- Cannot be answered using an observation
- Often ‘just’ based on agreement
- E.g.: ‘What do we mean with the ‘death penalty?’
3. Empirical (what is or what will be and why?)
- Can only be answered using observations. Conceptual and normative aspects are
relevant too, when asking empirical questions, but this is not what you hope to learn.
- 2 types of empirical research questions:
a. Descriptive questions
E.g.: ‘Which countries still have the death penalty?’ & ‘How many people were legally
executed between 2000 and 2005 in the USA?’.
b. Explanatory questions/causal
E.g.: ‘Does the death penalty reduce crime?’ & ‘Why did countries abolish the death
penalty, while other countries still have it?’.
- Many explanatory questions have 2 variables, however not all questions with 2 variables are
‘causal’, e.g.: ‘Do countries with a large Muslim population more frequently have the death
penalty than countries with other religious majorities? (= Descriptive question).
Pagina 2 van 22
, - Many explanatory questions do not have 2 variables, however not all questions refer
explicitly to 2 variables, e.g.: ‘Why do some countries abolish the death penalty, while other
countries still have it?’ (= Explanatory question).
Conformation Bias: looking for information that you already believe -> affects what we think is true.
3 types of conformation bias:
1. Information (looking for information that confirm our beliefs) --->
2. Reasoning (ignore some things, and stress the things that we believe in) --->
3. Conclusions (possibility to only remember the conclusions that are in line with our believes).
Reasons for conformation bias: emotional issues, convictions, sloppiness
Why is conformation bias so strong?
- Limitations (in humans)
- Wishful thinking
- Consistency (between initial evidence and new evidence)
Consequences of conformation bias:
- Mistakes in knowledge
- Bad decisions
Avoiding conformation bias:
- Not intelligence
- Systematic data collection (being aware of what can go wrong)
- Explicit and complete reasoning
Units, variables and Setting in Empirical Research Questions
Formulating clear ERQs by clarifying the Units of Analysis (object the research is about), the Variables
(possible characteristics/attributes of these units: values (e.g.: numbers, age, weight) and attributes
(e.g.: hair color, religions)) and the Setting (time and place/context).
E.g.:
- ‘Does the level of integration in society affect the amount of criminal behaviour among young
people in Western Societies?’
- ‘Does intercultural training affect the extent to which police officers discriminate against ethnic
minorities in Germany?
In many vague ERQs, the Units, Variables and Setting are unclear and cannot meaningfully be
identified, they need to be changed.
Research questions are about theoretical variables describing units.
Pagina 3 van 22