Chapter 1 - The Scientific Approach to Politics Flashcards
4 Examples of the Ubiquity of Social Science Research
- Public Opinion Data
- Behavioural Data
- Official Statistics
- Economic Data
Social Science Influences these 3 Things
- Political Debates
- Popular Beliefs
- Policy (sometimes)
3 Challenges for Consumers of Social Science Research
- Assessing the quality of studies
- Assessing the relevance of studies
- Ensuring objectivity
Definition of Evidence and the 2 Types
Information that is used to provide a basis for a point of view
- Facts
- Observations
2 Ways to Determine Values of Evidence
- Quality (validity, reliability)
2. The body of evidence
Argument Definition
Series of logical statements that lead to a conclusion, with reasons offered to support the conclusion.
Research Nihilism
That extreme end of critically evaluating research. There is no perfect research, but this belief posits that no research can have meaning because of inherent human bias and possible design flaws.
The 4 Parts of the Evidence Quality Continuum
Misleading Evidence
No or limited evidence
Best available evidence
Strong evidence
How to Distinguish and Approach Evidence
We examine how to distinguish between evidence that, despite its relatively minor limitations, merits guiding our practice versus more seriously flawed evidence that should be viewed more cautiously.
Four Portions of POLS 256 and what we will look at
- Scientific Method
- Positivism and its critics
- theory - Design Issues
- Ethics
- Concepts and measures
- Sampling
- Comparative case study - Quantitative Research Designs
- Surveys and official statistics
- Experiments
- Quantitative analysis - Qualitative Research Designs
- Textual analysis
- Interviews, focus groups, observation
- Qualitative analysis
5 Steps to Evidence-Based Practice and Policy
- Formulate Question
- Search for best evidence available
- Critically appraise evidence
- Select intervention based on appraisal of evidence, expertise, and context
- Monitor Progress
Type-of-Evidence Continuum (Three types)
- Opinion
- Assertion (some that is held to be true but that cannot or has not yet been demonstrated to be true)
- Fact
Why is social science science?
Because social scientists still use the same steps / scientific method and have similar philosophical assumptions about the world, reality, and causal relationships
The only difference is the objects of research
Normative v Empirical Analysis
Empirical is descriptive, explanatory, observation
Normative is prescriptive, ideals, value judgements
4 Underlying Assumptions of Positivism
- Causality / Determinism
- Measurable / Empiricism
- Objectivity
- Replication
Causality / Determinism Explained
Causal relationships exist, are the reason for the the state of the world.
Events are determined by previously existing causes.
Measurable / Empiricism
Things in the world are measurable through human perception / observation. Through this we are capable of understanding the causes of effects.
Objectivity
Despite being biased subjects, if we used certain methods we can objectively observe the single reality.
Replication
- Ideas and theories are continuously tested to ensure their place in reality. No amount of replication with the same result can lead to certainty, but one replication that disputes a claim ruins it forever.
- We can determine universal truths about the objective world by having many subjects use the scientific method to observe things.
4 Criticisms of Empirical Analysis / Evidence
- Evidence can change over time
- Evidence has its limits - humanistic element
- Evidence is filtered through power (postmodernism)
- Measurements - always imperfect, always things that are immeasurable
5 Challenges of Evidence-Based Policy
- Evidence searching, time consumption
- Studies may not exist
- Studies may point in different directions
- Studies may have poor design
- Intervention may not fit context
3 Challenges to Objectivity
Confirmation Bias
Disconfirmation Bias
Motivated Reasoning
- When you find evidence that should threaten your belief but you rationalize and find a way to make it fit)
Intersubjectivity
Empirical facts must be independently observed and agreed upon by many people.
(Two people, acting independently, perceive the same thing)
9 Step of the Scientific Method
- Identify the problem
- Hypothesize the cause of the problem
- Provide clear definitions of the concepts
- Operationalize the concepts
- Gather empirical data
- Test the hypothesis
- Reflect back on the theory
- Publicize the results
- Replicate the results
Empirical Analysis
Descriptive in nature; the goal is to describe and to explain the political as it is rather than as it should be.
Epistemology
An approach to knowledge
Independent Variable
The observable characteristics of phenomena that can take on more than one value; specific and concrete measurements of a concept
Epistemology
An approach to knowledge
Confounding
The presence of some unobserved difference between groups that is considered correlated both with the outcome and the independent variable of interest, thereby making causal inference impossible.
Treatments
The intervention of interest; also known as stimulus.
Postmodernism
Reality is socially constructed and the observer cannot be separated from what they purport to see.
Variance
In the context of the scientific method, there must be enough variance when the independent variable is removed to be measurable by human perception.
Interpretivism (Aka anti-positivism)
The view that reality does not exist independent of individuals and that all knowledge is socially constructed; no objective reality
Operationalizatoin
Moving from the general or conceptual level to the specific or concrete. Moving from concepts to variables.
In quantitative research, it would be assigning an abstract concept with a numerical value.
Postmodernism
A philosophical position that states that it is impossible to identify objective knowledge because all knowledge is embedded within the power, class, gender, and racial structure of society.
Correlation
The measurement or observation of a common variation among multiple concepts or measures; can exist without a causal relation.
Falsifiable
It must be possible, through both the logic of analysis and the design of research, to demonstrate the absence of a causal relationship between the concepts being examined.
Null Hypothesis
The opposite of the hypothesis - that there is no causal relationship between two variables. This must be tested in addition to the normal hypothesis.
Measurement Validity
The extent to which the measurement of a particular concept matches its operational definition.
