Lecture 19: Critical Thinking About Causality Flashcards

1
Q

What are Mill’s 3 criteria for establishing a causal relationship and what do they mean

A
  1. Priority = change of X precedes the change of Y
  2. Consistency = the change of X varies systematically which the change of Y
  3. Exclusivity = there is no alternative explanation for the relationship
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What are three errors in causal reasoning and explain what that means for the three criteria

A
  1. Mistaking correlation for cause = priority and exclusivity are not met
    X correlates with Y
    If X correlates with Y, then X is the cause of why
  2. Inversion of cause and effect = exclusivity is not met
    X causes Y
    If X causes Y, then no Y without X
  3. Post hoc ergo propter hoc = consistency is not met
    X precedes Y
    If X precedes Y, then X is the cause of Y
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is an INUS condition and what do the 4 factors stand for

A

It is an insufficient but non-redundant part of an unnecessary but sufficient condition

Insufficient = you need more than just the one thing to be able to achieve goal
Non-redundant = whether it’s there or not makes a difference
Unnecessary = you don’t need this one thing to achieve goal, you can also use other things
Sufficient = it is enough to achieve the goal

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is a counterfactual

A

A perfect counterfactual is knowledge of what would have happened to each participant if they had not undergone a certain manipulation —> physically impossible (can’t test people on something and also not test them)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What do we do to try and create a counterfactual, what is an issue with that and what is the design in which we do this called

A

Compare different groups of people, an issue is that people can have different baselines and the design is called a quasi-experimental design

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What are 7 threats to causal inference and explain them

A
  1. History: Influences (outside of intervention) over the course of the research, which influence
    outcome
  2. Maturation: Natural changes that may be confused with effect treatment
  3. Selection: Selection criteria for treatment related to outcomes of treatment
  4. Attrition: Participants’ dropout, systematically correlated with conditions
  5. Instrumentation: Change in measuring instrument resulting in a difference between pre- and post- measurement
  6. Testing: Effect of measurement itself on subsequent measurements (fatigue, habituation etc)
  7. Regression to the mean: Extreme scores will be followed by less extreme scores
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What are three ways in which people deal with the uncertainty between correlation and cause

A
  1. Ignoring causality (stating nothing about it)
  2. Statements of causality, but with unclear assumptions
  3. Pseudo-correlational statements (implying something about causality, but not actually stating it)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What are 3 causal diagrams when having 3 variables and what are they also called

A
  1. Mediation - meditator
  2. Common cause - confounder
  3. Common effect - collider
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is the purification principle and why is it correct/incorrect

A

The idea that the more control variables are included in a model, the more accurate the estimation of the causal effect is.
This is incorrect because it could lead to an underestimation of the total causal effect, or lead to collider bias

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is the collider bias

A

Controlling for common effects will bias the estimation of a causal relationship between two variables

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

When do we control for a third variable and when do we not

A

Control:
- when the third variable is a confounder
- when the third variable is a mediator AND you want to know the direct effect

Not control:
- when the third variable is a collider
- when the third variable is a mediator AND you want to know the total effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

T/F: if, in principle, all confounders are controlled for, a correlation between treatment and outcome can be seen as causal

A

True

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What are 2 benefits of randomized controlled trials

A
  1. They eliminate confounder bias; disables confounders
  2. Enables researcher to quantify uncertainty
How well did you know this?
1
Not at all
2
3
4
5
Perfectly