Pol 51 Midterm

Rows

An observation, one set of data

Columns

The variables

Independent variable

x or the cause

Dependent variable

y or the effect

Null Hypothesis

Hypothesis if 2 variables are unrelated. If unrelated, then the change in X will have no effect on the change in Y

Alternative Hypothesis

hypothesis that states what the expected relationship between X and Y will be (The hypothesis the researcher is trying to study)

Operationalization

Going from abstract concepts to empirical measures

Numerical variables

continuous, discrete

categorical variables

nominal, ordinal

continuous

relies on a specific ratio or equation such that you can get any value, unlike ordinal measures. The key distinction is the difference between units, which are always the same (infinite number of values between any two values)

• GDP, age, vote share, temperature

discrete

countable in a finite period of time, typically take on integer values

• Migrant deaths, number of students in the class, years of school

Nominal

discrete categories that a variable can fall into. No associated order or ranking

• Gender, religion

Ordinal

a scale with whole number values. We do not assume the difference between any two proximate pairs is the same, but treat them as if they were

• Education level, 5 points likert scale (strongly agree, agree, neither agree or disagree, disagree, strongly disagree)

Reliability

the extent to which the measurement technique yields the same results when repeated

Validity

the extent to which the measurement measures or captures the underlying concept.

Central tendency

typical values for a particular variable. MEAN, MEDIAN, MODE

Variation

the spread or range (dispersion) of values for a particular variable. Standard deviation

Mean

average

Mode

most common

Median

at the center of a variable range

skews

Data ___ to the outliers

Standard Deviation

• Broadly tells us how many observations fall within a certain “distance” of the mean.

• A measure of how dispersed the data is in relation to the mean

Fundamental Problem of Causal Inference (FPCI)

Cannont observe the counterfactual

Counterfactual

what would have happened if a specific condition or treatment did not happen

Explore causation

A scenario in a hypothetical world where one variable was actually different from what it was, allowing us to ___

Counterfactual ideal

we want to two parallel universe in which everything is identical except for X (controls for all confounders)