Lesson 1

Question 1

Why do we do statistics?

Answer 1

To systematize the way we account uncertainty when making data-based decisions.

Question 2

What can we say about the p-value in: Consider the one-tailed test for our estimated test-statistic of 𝑡Ƹ= 1.86 that produces a p-value of p = 0.032.

Answer 2

There is a 0.032 probability of observing a test statistic at least as large as 𝑡Ƹ, if the null hypothesis is true.

Question 3

What is the difference between Inference and Prediction?

Answer 3

• Inference focusses on the relation of variables to the outcome
• Prediction focusses on building a tool that can guess future values.

Question 4

What best represents a strength of statistical modeling relative to statistical testing?

Answer 4

The ability to control for confounding factors.

Question 5

True or False: A p-value is a test statistic

Answer 5

False

Question 6

How does the Data Science Cycle look?

Answer 6

1. Define problem
2. (Formulate hypotheses)
3. Collect data
4. Process data
5. Clean data
6. (EDA)
7. (Modeling/Testing)
8. (Evaluate)
9. (Report findings)
10. (Build data product)

Question 7

What is the difference between exploratory and confirmatory data analysis

Answer 7

When the data is well understood -> confirmatory.
If we don’t care about testing hypotheses -> exploratory.

EDA can be used to generate hypotheses for CDA and sanity check hypotheses.

Question 8

Why do we need statistical reasoning?

Answer 8

To quantify the uncertainty of our conclusions

Question 9

What defines a sampling distribution?

Answer 9

A mathematical function that describes all of the possible values that a statistic can take.