Political Science Stats Midterm

Statistics

The science of collecting, analyzing, presenting, and interpreting data.

Why is statistics so influential?

It is evidence-based and is one of the strongest bases of evidence.

Descriptive Statistics

Used to describe, summarize, and/or analyze collected data.

Inferential Statistics

Using a sample to draw a conclusion about the population.

Measures of Central Tendency

Statistical methods used to measure the average or central position for a single variable.

Public Policy

Laws, regulations, guidelines, or ordinances created to solve societal issues.

Cherry Picking

Only using data that supports your claim.

Correlation vs. Causation

Drawing conclusions without establishing a clear cause-and-effect relationship.

Small Sample Size

Generalizing based on a sample that is too small.

Conflict of Interest

Failing to disclose a potential conflict of interest and/or financial compensation.

Artificial Intelligence

A set of technologies that enable computers to simulate human intelligence.

Symbiotic Relationship between AI and data

AI requires massive amounts of data to learn and improve decision-making, while AI can also analyze massive amounts of data.

Unit

A single thing.

Data

A collection of facts, numbers, words, observations, or other useful information.

Measurement

Determining the quantity of qualities of something

Unit of Observation

The singular unit in a broader study.

Variable

The characteristic of the thing that we are studying.

Unit of Analysis

The thing being studied that frames what the study is about.

Conceptual Definition

The dictionary or textbook definition of a concept that helps us visualize what we are studying.

Operational Definition

The process by which we translate observations of reality into a measurement.

Validity

The measurement used for a concept makes logical sense.

Face Validity

A measure is considered valid because it makes sense.

Consensual Validity

A measure is considered valid because it has widespread use among social scientists.

Associational Validity

A measure is considered valid because it is correlated with other measures connected to the concept being studied.

Predictive Validity

A measure is considered valid if it is a good predictor of an effect we are trying to explain.

Reliability

The measure is consistent through time or across research.

What is reliability dependent on?

How well you define your measure.

Theory

A tentative conjecture about the causes of some phenomenon of interest.

Causality

A relationship between two things in which one causes the other.

Empirical

Evidence based on observing the real world.

Normative Statement

A statement about how the world ought to be.

Generality

Creating theories that can be applied to a general class of phenomena.

Why is generality important?

The less general a theory is, the less useful it is.

Parsimony

Creating theories that are simple and use minimal conjecture.

Why should theories be parsimonious?

Simpler theories are preferred.

Occam's Razor

Simpler theories are preferred over complex theories.

Independent Variable

The variable that is stable and unaffected by the other variable being measured; it is the presumed cause and is represented by X.

Dependent Variable

The variable that depends on other factors being measured; it is the presumed effect and is represented by Y.

What is another way to write a dependent variable?

Y-axis.

Spatial Dimensions

Identifies the unit of analysis from which we want to collect information.

Time Dimensions

Identify the point or points in time at which we want to measure a variable.

Cross-Sectional Study

The time dimension is the same for all cases and the dependent variable is measured for multiple spatial units.

Time-Series Study

The spatial dimension stays the same for all cases and the dependent variable is measured at multiple points in time.

Hypothesis

A theory-based statement about a relationship that we expect to observe, framed so that it can be empirically tested.

Null Hypothesis (H₀)

There is no effect or relationship between the independent and dependent variables.

Hₐ / Alternative Hypothesis

There is an effect or relationship between the independent and dependent variables.

Levels of Measurement

The numbers associated with the operationalization of a concept.

What are the four levels of measurement?

Ration, Interval, Ordinal, Nominal

Nominal Data

Data organized into categories with no natural order; it only identifies membership in groups.

Examples of Nominal Data

Race, gender, hair color, and astrological signs.

Ordinal Data

Data organized into categories that can be ordered from less to more, but the distance between categories is not uniform.

Examples of Ordinal Data

Class and educational degrees.

Interval Data

Data with an order and equal measurable distances between values, but without an absolute zero; it can contain negative values.

Ratio Data

Data with an absolute zero, meaning zero represents a total absence of the attribute being measured.

What types of variables are probably ratio level?

Percentages or physical measurements such as size, weight, quantity, or amount.

Coding

The process of translating information into numbers or character strings for analysis.

What are the four basic steps of coding?

Create a coding sheet, collect and record information, transfer the data to a database, and clean up the data.

Manual Coding

Coding that requires interpretation and judgment by the coder and may vary between coders.

Intercoder Reliability

When two different coders work on the same type of data, their results should match about 80% of the time.

Mean

The average; divide the sum of all values by the number of cases.

μ (Mu)

The mean of a distribution.

N

The total number of cases in a population.

Median

The middle score of a set of data.

Why is the median useful?

It is less affected by outliers and skewed data.

Mode

The most frequent score in a dataset or the most popular option.

What is the only measure of central tendency that can be reported for nominal-level data?

Mode.

What is the best measure of central tendency for nominal data?

Mode.

What is the best measure of central tendency for ordinal data?

Median.

What is the best measure of central tendency for interval/ratio data that is not skewed?

Mean.

What is the best measure of central tendency for interval/ratio data that is skewed?

Median.