Psc 310 final URI

ordinal variable

There is ranking order (liberal ——- conservative)

Times series data vs cross-sectional

Time series data: can be spread out in time Cross-sectional: observations from a certain point in time

Issues that comparative method faces

Too many variables, too few cases Selection bias

Control groups dont recieve treatment

True

Unit of analysis

Frames what is being analyzed as a whole (causality and change)

Bivariate analysis

Has low internal validity and high external validity because sample data is representative of whole population and cant rule out endogeniety or spurriousness

Experimental design typically has ___ levels of internal validity and ___ levels of external validity

High, low *because participants are sample of convenience

Measuremeant validity: face and content

Includes all aspects and elements of a concept Face: Appears to measure what its supposed to Content

Where does causual theory come from?

Induction

Casual theory: induction and deduction

Induction: observing seeking to find patterns pointing to theory Deduction: reasoning from general principals

Internal and external validity

Internal: judges how well research design tested a causal relationship (rulling out spurrious and endogeniety) External: confidence of generalizing finding to real world

Most similar system design

Key is to identify what facor leads to dissimilar outcomes of Y when the cases appear similar mostly

Most different system design

Key is to understand that different unit/cases have the same outcome/Y variable, then find explanatory variable common to the cases that all appear different from eachother

mean, median, mode

Mean: average Median: middle number after ranked Mode: most occuring

nominal variables

No ranking in between categories

Is random sampling required for experiments

No, random assignment is key

4 hurdles of causal theory

1) credible causal mechanism? 2) correlation does not equal causation 3) avoid Y causing X (endogeniety) 4) avoid Z causing X and Y (spurrious)

4 ways to develop good causal theory

1) offer answers 2) be non obvious 3) be nee 4) be general

Standard deviation

Captures how spread out/ dispearsed your data is *more spread =higher standard deviation

Qualitative case study

Costly to study a large number of cases

DV and IV

DV: phenomenon you want to explain (research question) IV: things that can explain phenomenon

Deterministic vs. Probabilistic

Deterministic: cause leads to effect with certainty Probablistic: cause leads to effect with increased probability

Qualitative

Explores causal mechanisms by researching a smaller number of cases

Continuous variable

Ranking order and equal unit differences between categories (age, income)

Normative statement

statement which describes how the world should be "we should have small governments"

measurement bias

the systematic over-reporting or under-reporting of values for a variable

Measurement Reliability

consistency of a measure

Ranking of importance

Valid, reliable, unbiased

Hypothesis

What we expect to observe based on our theory

null hypothesis

What we expect to observe if our theory is uncorrect (no connection/relationship between variables)