Quiz: Crash Course in Causality
- What is causality in data science?
a. The relationship between two variables where changes in one variable cause changes in the other variable.
b. The relationship between two variables where they both change in response to a third variable.
c. The relationship between two variables where they are independent of each other.
d. The relationship between two variables where changes in one variable have no effect on the other variable.
Answer a :The relationship between two variables where changes in one variable cause changes in the other variable.
2. What is the difference between correlation and causation?
a. Correlation is the relationship between two variables where changes in one variable cause changes in the other variable, while causation is the relationship between two variables where they both change in response to a third variable.
b. Correlation is the relationship between two variables where they are independent of each other, while causation is the relationship between two variables where changes in one variable have no effect on the other variable.
c. Correlation is a statistical relationship between two variables, while causation implies a direct relationship between two variables.
d. Correlation and causation are the same thing.
Answer: d) Association. While association is necessary for causality, it is not sufficient on its own to establish causality.
3. Which of the following is a common method for establishing causality in data science?
a. Random sampling
b. Statistical significance testing
c. Experimental design
d. Descriptive statistics
Answer: A confounding variable is an extraneous variable that is not part of the research question but can affect the relationship between the independent and dependent variables.
4. What is a confounding variable?
a. A variable that is affected by the independent variable and also affects the dependent variable.
b. A variable that is independent of the other variables in the study.
c. A variable that has no effect on the dependent variable.
d. A variable that is not included in the study.
Answer a. A variable that is affected by the independent variable and also affects the dependent variable.
5. What is the counterfactual approach to causality?
a. It involves comparing the outcome of an event to what would have happened if the event had not occurred.
b. It involves manipulating the independent variable and observing the effect on the dependent variable.
c. It involves randomly assigning participants to different groups in an experiment.
d. It involves measuring the correlation between two variables.
Answer: A counterfactual is a hypothetical scenario in which an event did not occur, allowing researchers to compare the actual scenario with a hypothetical one to identify the causal effect of the event.
6. Which of the following statements is true about the Bradford Hill criteria for causality?
a. They are a set of statistical tests used to determine causality.
b. They are a set of guidelines for assessing the strength of evidence for causality.
c. They are a set of experiments used to establish causality.
d. They are a set of observational studies used to establish causality.
Answer a. To identify and measure causal effects.
7. What is the difference between direct and indirect causality?
a. Direct causality is the relationship between two variables where one variable causes the other variable, while indirect causality is the relationship between two variables where they both change in response to a third variable.
b. Direct causality is a statistical relationship between two variables, while indirect causality implies a direct relationship between two variables.
c. Direct causality and indirect causality are the same thing.
d. Direct causality is a type of confounding variable.
d. All of the above methods can be used to estimate the average causal effect.
8. Which of the following statements is true about the potential outcomes framework for causality?
a. It involves comparing the outcome of an event to what would have happened if the event had not occurred.
b. It involves manipulating the independent variable and observing the effect on the dependent variable.
c. It involves randomly assigning participants to different groups in an experiment.
d. It involves measuring the correlation between two variables.
Answer b. An RCT involves random assignment of participants to treatment and control groups, while an observational study does not.
9. What is the difference between an observational study and an experiment?
a. Observational studies involve manipulating the independent variable, while experiments do not.
b. Observational studies involve randomly assigning participants to different groups, while experiments do not.
c. Observational studies do not involve manipulating the independent variable, while experiments do.
d. Observational studies and experiments are the same thing.
e. None of the above
Answer e. None of the given options is a common assumption of causal inference. The common assumption of causal inference is that there are no unmeasured confounders or hidden variables that could explain the relationship between the independent variable and the dependent variable.
10. Which of the following is NOT a method for estimating causal effects?
a. Difference-in-differences
b. Randomized controlled trials
c. Principal component analysis
d. Instrumental variable analysis
Answer: c. Principal component analysis
