Week 1: Research, theories, research goals, hypotheses, and concepts
Research
- Research = providing answers to questions we don’t know answers to. Happens with
open mind (not pursuing confirmation of pre-existing belief) to provide valid, precise
and new answers.
- Design of research ensures that the answers are as valid as possible and discovered
efficiently.
- Steps: problem definition theory development conceptualization
operationalization variable selection case selection
Aspects of science
1. Publicness: research public + open to scrutiny. Allows for gradual improvements and
correction of mistakes by replicating, adjusting, and overturning studies. Academic
research mostly publicly funded and contributes to common good instead of private
gain.
2. Scientific method: adherence to rules regarding theory development, measurement,
inference from data, and other aspects of the research process.
Dividing lines in science
- Positivists vs. subjectivists
- Empiricists vs. scientific realists
- Qualitative vs. quantitative
Disagreement: subjectivists vs. positivists
- Disagree whether social reality can be studied scientifically.
- Subjectivists: social science is not value-free. It’s influenced by subjective nature of
human perception and experience. Research must focus on interpreting the meaning of
and reflecting on reasons for human action.
- Positivists: argue that social reality is inter-subjective. However, the causes and
mechanisms of social phenomena can be discovered using scientific rules, standards,
and procedures.
Disagreement: empiricists vs. scientific realists
- Disagree about the ways to conduct study. Imply different ontological views (what is
reality) and differences in epistemology (how to know reality).
- Empiricist: use instrumentalist view of theoretical assumptions (assumption of theories
and models don’t need to be realistic as long as they work, i.e., prove useful for
prediction and manipulation). Only interested in successful prediction and intervention.
- Scientific realists: say that studies have to provide deep understanding of social reality.
Disagreement: qualitative vs. quantitative
- Quantitative: focus on numbers and statistics
- Qualitative: focus on deep case studies.
Conspiracy theories unscientific because they are not falsifiable.
1
,Limitations human decision-making
1. Confirmation bias: seeking only information that confirms a hypothesis but no
information that would contradict it.
2. Hindsight bias: making believe that events have been more foreseeable than they
actually were
3. Framing effect: leading people to make different inferences from same information by
change of words
4. Availability bias: focusing on most salient information that comes first to mind
Types of research: normative vs. positive
- Normative: is prescriptive (what ought to be) and based on norms and values.
Normative issues examine the moral order of the world and seek to find answers to
what is good, appropriate, just, fair or desirable. Can’t be answered with empirical data.
- Positive: is descriptive (what is) and about relationships between concepts and empirical
facts. Focused on describing, understanding, explaining or predicting reality as it is. Two
kinds:
1. Purely theoretical (rare): studies relationships between theoretical concepts without
direct reference to empirical facts.
2. Empirical positive research (usual): connecting theoretical concepts to empirical
facts. Goals of empirical research are description, prediction, and explanation.
- Distinction between normative and positive is important because helps separate
subjective value judgements from objective statements about facts or concepts. Helps to
enter the research process with an open mind. However, positive research is usually
motivated by normative concerns and contribute to ethical discussions.
Theoretical research
- The goal is to derive a set of statements (propositions) that are internally consistent and
logically related with each other.
- Theoretical statements are analytic: truth of analytic statements derived only and
entirely from their premises and rules of logic. If premises are true, conclusions are true.
- Theory (facts and ideas) is generated from the empirical world and the assumptions are
then tested on the empirical world.
Empirical research
- Empirical research is any research that makes reference to real-life phenomena and
empirical facts.
- The way of thought is that some form of knowable ‘reality’ exists, and this reality can
and needs to be described, interpreted, and explained. Statements with empirical
content are synthetic: theory plays a role, but truth relies on real-life phenomena and
empirical facts
Types of research
- Normative and positive. Within positive, there is theoretical and empirical research. And
within empirical research there is descriptive, predictive and explanatory research.
2
, Goals of empirical research
1. Descriptive research: tries to describe to world as it is. It assigns facts into classes or
categories that can be used in theory building or hypothesis testing. Description
relies on inference (= observing something and concluding something about
something else. Inference from sample to population).
2. Predictive research: projecting empirical phenomena and processes to the future.
Sometimes, even if we have good explanatory theories, prediction can’t be used
because:
1. We can only arrive at probabilistic predictions.
2. Sometimes we can understand causal mechanisms relatively well, but still remain
impotent to predict due to nature of processes producing the events.
3. Might be unable to predicts events in the future if those factors are not
predictable or possible to measure.
3. Explanatory research: focus on causes of events, identifying general causal effects,
and revealing the causal mechanisms that produce them. The goals can be to explain
a particular case, to build a comprehensive causal account of a general
phenomenon, or to establish a general causal relationship for a population of cases.
Two kinds of explanations:
1. Complete explanations: build a comprehensive causal model of a
phenomenon
2. Partial explanations: focus on a single possible causal relationship
Difference with the other two: causal explanations offer deep understanding and
possibility for manipulation and control. Explaining means understanding the
causal structure which implies the ability to manipulate this structure and change
the outcomes.
Inference
- Making valid statements about the whole from observing only a part in order to
generalize outcomes.
- All research requires inference, even descriptive and predictive. Explanatory research
requires special type of inference, called causal inference.
What type of empirical research is used for lesson-drawing?
- Lesson-drawing requires double causal inference policy recommendations require
normative research on top of explanatory research on top of descriptive research.
Two other types of research
1. Explorative research: not a distinct research goal; it is an advanced form of description,
but in a more unstructured way. Goal is to explore if something is different for other
populations. Still descriptive, no causal statements.
2. Interpretative research: more explanatory. Focus on subjective understanding of people.
Explanatory, but not about causal explanations.
3
