Module 2: Evaluating Models

theory

A general set of explanatory claims about some specifiable empirical range.

hypothesis

A specific prediction, derived from a theory, that can be tested against empirical evidence.

deductive reasoning

The process of moving from general claims or theories to specific observations or predictions about a phenomenon or set of cases.

deviant case

A case that does not fit the pattern predicted by a given theory.

inductive reasoning

The process of moving from specific observations to general claims.

thesis

A statement for which one argues on the basis of evidence.

qualitative

A form of analysis that aims to discern relationships between events or phenomena as described in narrative form, such as an account of an historical process.

quantitative

Quantitative analysis aims for the mathematical discernment of relationships between variables, typically involving a large number of cases or observations.

inference

The process through which we aim to test observable implication (often about cause and effect) of any given theory.

correlation

A relationship between two variables in which they tend to move in either the same direction (positive correlation) or in opposite directions (negative correlation).

causation

The property that obtains when one thing can be shown to cause another.

falsifiability

The testability of a theory or hypothesis. A good hypothesis could be logically demonstrated to be false by evidence.

endogeneity

The name given to any circumstance in which two variables exhibit mutual or reciprocal effects.

empirical critique

An effort to point to important evidence that does not support a conventional version of any given theory.

theoretical critique

An effort to show that a given theory has logical limitations.

scope conditions

The conditions or range of cases for which an argument works.

methodological critique

An effort to point to important problems with the way variables are measured or hypotheses are tested

indicator

An element or feature that indicates the presence of an underlying factor.

measurement error

Either an episodic error, such as improperly recording data, or a systematic error, meaning that a measurement does not fully reflect what it is designed to measure.

measurement bias

A measure is biased if it will not produce comparable results for all observations.

measurement validity

Whether a given measure effectively captures or represents what we are researching.

Truth

a criterion for assessing a causal model's accuracy.

Circular model

a model that is not testable.

Critical experiments

a method of comparing alternative models by finding a question for which they give different answers.

Simplicity

a term used to describe a causal model that has a small number of assumptions is more attractive than one having a large number of assumptions

Fertility

a term used to describe a causal model that produces a relatively large number of interesting predictions per assumption.

Justice

a criterion for assessing a model's likely implications for the world.

Beauty

a term used to describe a model that is simple, fertile, and/or surprising

Causal process

a series of actions or steps taken that bring about a particular outcome.

predictor

a variable that predicts that something will happen in the future or will be a consequence of something; synonym for independent variable; indicator of independent variable

Institution

The humanly devised constraints that structure political, economic, and social interaction.

Modernization theory

A theory asserting that democracy can be expected as a consequence of the emergence of economic development, industrialization, education, and urbanization, all of which contribute to the emergence of a middle class.

Formal institutions

Institutions usually based on officially sanctioned rules that are relatively clear

Informal institutions

Institutions with unwritten and unofficial rules

Causal relationship

A connection between two entities that occurs because one produces, or brings about, the other with complete or great regularity.

Selection bias

Choosing cases with a focus on effects rather than causes, which can lead to inaccurate conclusions about correlation or causation

Rationalist

theoretical approach that assumes that individuals weigh the costs and benefits and make choices to maximize their benefits

Culturalist

theoretical approach that asks how rules shape individual and group identities

Structuralist

theoretical approach that asks how actors relate to each other through institutions