ISYE 6414 Midterm Exam: Quiz |Questions and Answers
Latest Updated 2026 (Graded A+) - Georgia Institute Of
Technology
Due Jun 16 at 11:59pm
Points 40
Questions 25
Available Jun 9 at 8am - Jun 16 at 11:59pm
Time Limit 60 Minutes
Allowed Attempts Unlimited
Instructions
This section is open-book but NOT open to the internet except StackOverflow. That means not using
Google, StackExchange, ChatGPT, etc. Please do not search stack overflow on google. Go directly
to stackoverflow.com. You may use your written notes, files, lecture slides, and code examples for this
section of the exam. You may not communicate or receive help from anyone until you have submitted
your exam. You may use your computer's onscreen calculator app or a physical calculator but may not
use your phone/tablet calculator. Phones/tablets are prohibited in both exam sections.
It is your responsibility to keep track of your time and submit before the time limit.
The Honorlock support team is available 24/7. The Honorlock support agents provide best-in-class
support, and each one is trained to offer quick and consistent assistance. Whether you’re testing at two
in the morning or on a weekend evening or over a holiday, you can call at any time and get help from a
human. They'll troubleshoot anything you’re having problems with and make your online testing
experience as smooth as possible.
Honorlock Student Support: 1 (844) 243-2500
Honorlock Student Support Email: (mailto:sup )
Honorlock Student Hours: Support is available 24/7/365
1/15
, Take the Quiz Again
Attempt History
Attempt Time Score
KEPT Attempt 2 12 minutes 36 out of 40
LATEST Attempt 2 12 minutes 36 out of 40
Attempt 1 23 minutes 29.67 out of 40
Submitted Jun 13 at 4pm
Question 1
pts
How would we interpret Cook’s Distance of outliers for a large sample size (n>10,000)? (Hint: an
example was given in the lecture notes for Bike Rental Data)
We can split the data into batches and remove outliers with 4/n – k for each of the k batches.
Use a hard rule: Eliminate all points with a Cook’s Distance greater than 4/n.
Correct!
Look at the Cook's distances holistically even if some of them are larger than the threshold of 4/n.
Use a hard rule: Eliminate all data points less than 4/n.
Module 2 Lesson 2.13 - slide 8
Our rule of thumb, 4/n, does not hold for large n. We need to look at the Cook’s Distance in the lens of
the dataset. The values may end up being small and the 4/n is not always a hard rule to follow.
Question 2
pts
Which statement is true regarding the "Inflated Statistical Significance"?
It consists of building multiple linear regression models and adjusting the predictors to get the lowest p-values.
It’s the problem of drawing conclusions on statistical significance based on p-values and ignoring residual analysis.
Correct!
It can be resolved by sub-sampling and drawing conclusions from resulting empirical distributions of the regression
coefficients.
It can be resolved by outlier elimination and reducing the threshold for predictor selection.
2/15
Latest Updated 2026 (Graded A+) - Georgia Institute Of
Technology
Due Jun 16 at 11:59pm
Points 40
Questions 25
Available Jun 9 at 8am - Jun 16 at 11:59pm
Time Limit 60 Minutes
Allowed Attempts Unlimited
Instructions
This section is open-book but NOT open to the internet except StackOverflow. That means not using
Google, StackExchange, ChatGPT, etc. Please do not search stack overflow on google. Go directly
to stackoverflow.com. You may use your written notes, files, lecture slides, and code examples for this
section of the exam. You may not communicate or receive help from anyone until you have submitted
your exam. You may use your computer's onscreen calculator app or a physical calculator but may not
use your phone/tablet calculator. Phones/tablets are prohibited in both exam sections.
It is your responsibility to keep track of your time and submit before the time limit.
The Honorlock support team is available 24/7. The Honorlock support agents provide best-in-class
support, and each one is trained to offer quick and consistent assistance. Whether you’re testing at two
in the morning or on a weekend evening or over a holiday, you can call at any time and get help from a
human. They'll troubleshoot anything you’re having problems with and make your online testing
experience as smooth as possible.
Honorlock Student Support: 1 (844) 243-2500
Honorlock Student Support Email: (mailto:sup )
Honorlock Student Hours: Support is available 24/7/365
1/15
, Take the Quiz Again
Attempt History
Attempt Time Score
KEPT Attempt 2 12 minutes 36 out of 40
LATEST Attempt 2 12 minutes 36 out of 40
Attempt 1 23 minutes 29.67 out of 40
Submitted Jun 13 at 4pm
Question 1
pts
How would we interpret Cook’s Distance of outliers for a large sample size (n>10,000)? (Hint: an
example was given in the lecture notes for Bike Rental Data)
We can split the data into batches and remove outliers with 4/n – k for each of the k batches.
Use a hard rule: Eliminate all points with a Cook’s Distance greater than 4/n.
Correct!
Look at the Cook's distances holistically even if some of them are larger than the threshold of 4/n.
Use a hard rule: Eliminate all data points less than 4/n.
Module 2 Lesson 2.13 - slide 8
Our rule of thumb, 4/n, does not hold for large n. We need to look at the Cook’s Distance in the lens of
the dataset. The values may end up being small and the 4/n is not always a hard rule to follow.
Question 2
pts
Which statement is true regarding the "Inflated Statistical Significance"?
It consists of building multiple linear regression models and adjusting the predictors to get the lowest p-values.
It’s the problem of drawing conclusions on statistical significance based on p-values and ignoring residual analysis.
Correct!
It can be resolved by sub-sampling and drawing conclusions from resulting empirical distributions of the regression
coefficients.
It can be resolved by outlier elimination and reducing the threshold for predictor selection.
2/15
