poli 30d midterm

4 research designs

randomized experiment → researcher randomly assigns qualities, high internal + low external

natural experiment → nature randomly assigns control + treatment groups, high internal + low external

quasi experiment → observe groups of cases with an artificial line, low internal + high external

controlled comparison → researcher observes effect of IV on DV in nature, low internal + high external

causality

capacity of 1 variable to influence another, bringing it into existence or causing it to fluctuate

4 criteria for causality

temporal ordering → IV before DV

correlation → changes in 1 variable occur w/ changes in the other

causal mechanism → plausible story that connects IV to DV

rule out confounds → rule out variable that may correlate to IV and DV and be the true cause

Popper

essence of scientific claim/theory is falsifiability, falsify probabilistic theories for social science

Comte

aim to discover natural laws and hierarchy of science, social physics are laws of society and behavior, social science is hardest to define

Kuhn

define normal science, operate within systems of paradigms, paradigms replaced when anomalies appear, social science is pre-paradigmatic

Merton

ideal science CUDOS, reality of social science differs

C → communalism, knowledge belongs to all

U → universalism, judged on evidence not identity

D → disinterestedness, pursuit of truth not gain

OS → organized skepticism, constant scrutiny

conceptual definition

unambiguous statement that defines the features of a concept, its variation, the unit to which it applies

operational definition

definition that makes research replicable, other scholar could use this definition and get the same results

confounders

variable that affects IV and DV and proves relationship is correlation, not causation

population

group/individual we want to analyze and say something about

population parameter

characteristic of the population

sample

subset of population for which we have data

sample statistic

characteristic of the sample

sampling frame

broader group you’re sampling

variable types

nominal → unordered, categorical variable

ordinal → ordered, categorical variable

interval → lacks true zero, numerical variable

ratio → has true zero, numerical variable

histogram

shows frequency distribution of single continuous numerical dataset, numerical X + Y axis

bar chart

represents frequency of nominal data, numerical Y axis, categorical X axis

validity types

internal validity → did IV cause DV, rule out confounds

external validity → does this tell us something about how things work in the real world

construct validity → does this tool truly measure the concept

face validity → does this tool appear to measure the concept

counterfactuals

hypothetical scenario describing what would happen if IV were different

crosstab

relationship between nominal/ordinal variables, calculate % based off IV