Lecture 2 & 3 - correlation vs causation & Experiments Flashcards
What is correlation?
Statistical relationship between two variables.
What is causation?
One event is the result of the other event, i.e. one caused an effect.
What are some causal terms associated with causation?
- leads to * brings about * impacts * reduces * increases * weakens
Why is it important to determine if X causes Y?
To understand relationships and predict outcomes.
What does it mean when correlation does not imply causation?
Correlation could be due to chance, reverse causation, confounding variables, or selection bias.
What is reverse causation?
A situation where the correlation suggests that Y causes X instead of X causing Y.
What are confounding variables?
A third variable that affects both the independent and dependent variables.
What is selection bias?
When the sample surveyed is not representative of the population due to a specific characteristic affecting participation.
What is the challenge of confounding in research?
Confounders are often hidden or unknown.
What is an example of confounding in exercise and mood?
Stress can confound the relationship between exercise and mood.
What should researchers be primarily concerned about when interpreting survey results?
Selection bias and confounding variables.
What does correlation indicate?
That something interesting might be going on, warranting further investigation.
What are two methods to distinguish correlation from causation?
- Use control variables in statistical analysis * Conduct experiments
What is regression analysis?
A statistical method used to identify correlation between independent and dependent variables while controlling for confounders.
What is the gold standard for eliminating confounding variables?
Randomized Controlled Trials (RCT).
What are the three key elements of a Randomized Controlled Trial (RCT)?
- Randomized allocation of participants * Treatment of one variable * Controlled setting
What is the purpose of randomization in RCTs?
To ensure that treatment and control groups are statistically equivalent.
What is the counterfactual in research?
What would have happened if a different decision had been made.
Fill in the blank: The relationship between diet soda and obesity may be an example of _______.
[confounding variables]
True or False: Correlation always implies causation.
False
What is the impact of controlling for confounders in statistical analysis?
It can increase the likelihood of finding causal relationships.
What is the challenge with media stories regarding correlation and causation?
Correlation is often exchanged with causality.
What does RCT stand for?
Randomized Controlled Trial.
What is deterministic causality?
A type of causality where the effect is guaranteed if the cause occurs.
