Block 1: X-centered RDs: Causal strategies: X and Y (Lesson 3)

Question 1

What is the golden standard for x-centered research?

Answer 1

The experimental template

Outcome (Y) is less important, the focus is on the effect on the population

Any deviations from this template are viewed as sources for bias

Question 2

Definition of ontology

Answer 2

The nature of being, existence or reality

Question 3

Definition epistemology

Answer 3

Theory of knowledge: It’s nature and the limitations thereof

Question 4

Definition of ethics/ aesthetics

Answer 4

What value does an observation have?

Question 5

Different types of causal relationships

Answer 5

CEMS LICS PCPCP

Conjunctures
Equifinality
Monotonicity
Sequence

Linearity
Irreversibility
Constancy
Set-theoretic causes

Proximal
Causal chain
Path-dependency
Causal laws
Probibalistic causes

Question 6

Conjunctures (causal relationship)

Answer 6

A combination of causes that produce an effect

Question 7

Equifinality

Answer 7

Several causes acting independently of each other, but lead to the same effect/ outcome

Question 8

Monotonicity

Answer 8

Where an increase (decrease) in X always causes an increase (decrease) in Y

Question 9

Linearity

Answer 9

Rise in X causes a predictable rise in Y, explainable by a linear relationship

Question 10

Irreversibility

Answer 10

One way relationship

X affects Y as X increases but not as it decreases (or vice versa)

Question 11

Constancy

Answer 11

A constant cause operates continually upon an outcome

Question 12

Proximal

Answer 12

A proximal cause operates immediately

Question 13

Sequence

Answer 13

The effect of X(1-3) on Y depends upon the sequence they appear in

Question 14

Causal chain

Answer 14

Multiple factors (M) form a chain between X and Y

Question 15

Path-dependency

Answer 15

A single causal intervention has enduring, and perhaps increasing, effects over time

Question 16

Causal laws

Answer 16

Exception-less relationships between X and Y

Question 17

Probabilistic causes

Answer 17

Relationship with errors, i.e., exceptions, which can be given a certain probability of occurring

Question 18

Set-theoretic causes

Answer 18

Where X is necessary and/or sufficient for Y

Question 19

Criteria for good causal analysis on the treatment variable (X)

p
e
e
v
s

s
u
d
s

Answer 19

Is X 
P =  proximate to Y,
E = exogenous to Y, 
E = evenly distributed,
V = varying, 
S = simple, 

S = strong, 
U = uniform, 
D = discrete, 
S = scaleable?

Question 20

Criteria for good causal analysis on the outcome variable (Y)

Answer 20

Is Y free to vary?

Question 21

Criteria for good causal analysis on the sample

Answer 21

Are the chosen observations (a) independent (of one another) and
(b) causally comparable?

Question 22

Definition of independence (sample criteria)

Answer 22

Each observation is seperate and gives new evidence of the causal linkage

Changes in Y need to be due to the treatment, not because the units themselves are influncing each other

Question 23

Definition of causal comparability (sample criteria)

Answer 23

The average value of Y for a given value of X should remain the same across units and during the period of analysis

Question 24

Incomparabilities which may influence causal comparability

Answer 24

Noise (B), which is random and only influences Y

Confounders (C), non-random and influence both X and Y

Question 25

Strategies for causal inference

Answer 25

Randomized designs (experimental) Non-randomized designs (non-experimental, observational) Beyond X and Y

Question 26

Randomized designs (strategies)

Answer 26

– Pre-test/post-test, – Post-test only, – Multiple post- tests, – Roll-out, – Crossover, – Solomon four-group, – Factorial

Question 27

Non-randomized designs (strategies)

Answer 27

– Regression discontinuity, – Panel, – Cross-sectional, – Longitudinal

Question 28

Beyond X and Y (strategies)

Answer 28

– Conditioning confounders, – Instrumental variables, – Mechanisms, – Alternate outcomes, – Causal heterogeneity, – Rival hypotheses, – Robustness tests, – Causal reasoning

Question 29

Generally on randomized designs and their internal/ external validity

Answer 29

Internal validity and the assignment problem are per definition solved, though post-treatmen threats can be present External validity can still be problematic

Question 30

Give two examples of pre-treatment/ assignment bias?

Answer 30

1) Common-cause confounder: Confounder C affects both the treatment (X) and outcome (Y) 2) Self-selection bias: Treatment assignment is done by the subjects of the study => Solved through randomization!

Question 31

Post-Treatment Bias, examples

Answer 31

Attrition: The loss of subjects during the course of a study Noncompliance: When subjects do not comply with instructions Contamination: Where treatment and control groups are not isolated from one another Reputation effects: Where the reputation of the treatment in the minds of subjects affects an outcome Researcher (Hawthorne) effects: Where the condition of being studied affects an outcome Testing effects: Where responses to a test are influenced by a previous testing experiences, rather than the treatment itself

Question 32

Examples of Pre/Post-Treatment Bias in Longitudinal Studies

Answer 32

– History: Where the treatment is correlated with some other factor that affects the outcome, which is to say, where the variation over time is driven by some factor other than the treatment. – Regression to the mean: Where some change observed over time is a product of stochastic variation rather than the treatment of interest. – Instrumentation effects: A change in the measurement of an outcome over the course of a study which alters the estimate of X’s effect on Y (learning effects). => Remain present even when randomization is used!

Question 33

Stochastic variable

Answer 33

Variable which is completely random with no systemic component (statistics)

Question 34

Generally on Non-Randomized Designs

Answer 34

– When randomization isn't possible, the assignment problem has to be dealt with – Except in case of (perfect) natural experiments, non- randomization is the standard in studies using observational data – The fundamental problem: We do not know the ‘data generating process’ (DGP), because we did not create the data ourselves

Question 35

Cartwright's socio-economic machine argument

Answer 35

We cannot compare countries because they have different socio-economic backgrounds and we know little about the data-generation process

Question 36

Non-randomized designs: Regression-Discontinuity Design (RDD)

Answer 36

Assignment principle: (a) known; (b) measurable (prior to treatment, for all units in the sample); (c) Has a cutoff point, which defines the assignment of subjects, producing a binary treatment variable; (d) many units fall on either side of this cutoff; (e) this principle is maintained throughout the period of study

Question 37

Non-randomized designs: Panel Design

Answer 37

Multiple observations are taken from each unit and there variation in X across time/ units. Two types: 1) Difference-in-difference (DD) 2) Fixed effect

Question 38

Difference-in-difference (DD) panel design formula

Answer 38

Y = B + T + X + T*X ``` B = covariates T = time dummy X = treatment ```

Question 39

Types of autocorrelation

Answer 39

1) Spatial autocorrelation: Non-independence because observations are correlated across space (neighborhood effect) 2) Serial autocorrelation: Non-independence because observations are correlated across time (temporal stickiness)

Question 40

Non-randomized designs: Cross-Sectional Designs

Answer 40

Based on post-test observations, where X and Y vary spatially (but not over time) Advantage: Creates fewer statistical (method) problems

Question 41

Non-randomized designs: Longitudinal Designs

Answer 41

Variation in variables is longitudinal (over time) but not cross-sectional, thus all units are treated Treatment effect is found by comparing pre-treatment and post-treatment status

Question 42

Sub-types of the longitudinal design

Answer 42

1) Interrupted time-series: Single intervention affects the units, which are observed ofer time, before and after the intervention 2) Repeated observations: Units receive the same treatment multiple times