wayground logo

Free Printable Worksheets

NEW

Font size

S
M
L
XL
Worksheets

M235 Module 5 Trivia

Total questions: 10

Worksheet time: 10mins

Name
Class
Date
1.

What can we say about the two key assumptions we make when performing causal inference with observational data?

a)

Overlap is an assumption that we can assess by looking at the data

b)

Unconfoundedness is an assumption that we can test for

c)

Balance of covariates is a key assumption necessary for inference

d)

All of the above

2.

What assumptions does sensitivity analysis attempt to examine when we are performing causal inference with an observational study?

a)

Tests for whether the unconfoundedness assumption is violated or not

b)

Assesses how sensitive the results are in regards to how much overlap there is

c)

Evaluates how sensitive the results are when the unconfoundedness assumption fails in certain ways

d)

Attempts to balance the covariates for more efficient inference

3.

What is not one of the four sensitivity parameters that Rosenbaum and Rubin vary in their model-based sensitivity analysis? Let U = one unmeasured binary confounder.

a)

Effect of U on treatment probability

b)

The probability that U = 1

c)

Effect of U on the outcome

d)

Weights on each observation

4.

How does Rosenbaum’s bounds approach to sensitivity analysis use less assumptions than Rosenbaum and Rubin’s model-based sensitivity analysis?

a)

It avoids making assumptions on how the unmeasured confounder affects the outcome

b)

It avoids looking at the association of the unmeasured confounder and treatment

c)

It uses randomization inference framework to avoid distributional assumptions on the outcome

d)

All the above

5.

Which of the following is not part of the three-factor decomposition in Tukey’s factorization when using Bayesian sensitivity analysis?

a)

The observed data distribution

b)

The propensity score model

c)

The marginal distribution of the observed outcomes

d)

The selection function

6.

What is the E-value?

a)

The minimum strength of association required for a confounder to explain away an observed treatment effect

b)

The probability of the unmeasured binary confounder being one

c)

The amount of bias in our treatment effect that results from having an unmeasured binary confounder

d)

The model-based estimate of how much we want to set the association between the unmeasured confounder and treatment probability

7.

What justification is there for assuming there is only one unmeasured confounder that is binary in model-based sensitivity analyses?

a)

Even if there are more than one unmeasured confounder, you can consider U to be a scalar summary of the other confounders (like propensity scores)

b)

Causal conclusions are more sensitive to unobserved binary covariates than continuous ones

c)

Sensitivity analysis is already a secondary analysis, so overcomplicating it with additional assumptions may distract from the main analysis

d)

All of the above

8.

When using causal trees, which variables should be used for splitting when constructing the tree structure?

a)

Covariates, X

b)

Outcomes, Y

c)

Treatment indicator, W

d)

All of the above

9.

What is not an advantage of causal trees?

a)

Good at filtering out noise, so less prone to overfitting

b)

Can handle non-linearity of covariates

c)

Automatically captures covariate interactions

d)

Can easily separate tree construction from treatment effect estimation

10.

When might you want to use causal forests over causal trees?

a)

When you want to be able to identify clear and interpretable subgroups

b)

When you want to directly estimate CATE as a function of the covariates

c)

When you want to estimate the CATE for specific, pre-specified subgroups

d)

When you have high dimensions and don’t want your CATE estimates to be biased