Section 1: Lecture 7 - Predicting Binary Outcomes

Section 1: Lecture 7 - Predicting Binary Outcomes

Created by
@JudiciousNephrite2042

Questions and Answers

What is the main goal when adding parameters to a model?

To get the most bang for our buck from a model's parameters

What happens to the prediction or generalization to new data when adding more parameters to a model?

It often gets worse

What is the starting point when building a model?

The saturated model

What is the purpose of removing the highest order interaction in the saturated model?

<p>To test the significance of the interaction</p> Signup and view all the answers

What is true about the likelihood ratio statistic (G2)?

<p>It has an approximated chi-squared distribution</p> Signup and view all the answers

What is the condition for removing 2-way interactions in a model?

<p>The change in fit is non-significant</p> Signup and view all the answers

What is the minimum number of cases per cell required for a good model?

<p>5 times as many cases as there are cells</p> Signup and view all the answers

What should be true about the expected cell frequencies?

<p>All should be greater than 1</p> Signup and view all the answers

What should be true about the standardized residuals?

<p>They should be normally distributed</p> Signup and view all the answers

What is the purpose of plotting standardized residuals against observed values?

<p>To identify patterns in the residuals</p> Signup and view all the answers

More Quizzes Like This

Use Quizgecko on...
Browser
Browser