Exam 3 review

STA 210 - Summer 2022

Yunran Chen



  • Exam 3 Review
  • Course Evaluation
  • Peer Review

Exam instructions

  • The exam is an individual assignment. Everything in your repository is for your eyes only.

  • You may not collaborate or communicate anything about this exam to anyone except the instructor. For example, you may not communicate with other students, the TAs, or post/solicit help on the internet, email or via any other method of communication.

  • The exam is open-book, open-note, so you may use any materials from class as you take the exam.

  • If you have questions, email me.

Exam coverage and format

  • Focuses on content after exam 2, but can include material from previous weeks

  • Similar format as previous exams

    • Part 1: Multiple choice/fill-in-the-blank questions on Sakai
    • Part 2: Open-ended data analysis in GitHub and submitted on Gradescope

Part 2 of the exam

  • Goal: Assess your understanding of the course material and how the methods you learned are applied to the analysis of real-world data.

  • Include all of your analysis steps in your exam write up, unless stated otherwise.

    • For example, if the exam says “assume conditions are met,” You can reference that information in your write up but don’t have to recheck the conditions.

Assessment criteria

  • You can identify the correct approach, analysis method, and/or inferential results required to answer the question.
  • You understand the correct conditions and diagnostics needed to determine whether the conclusions drawn from the model will be reliable
  • You can write results and conclusions in a meaningful way that can be understood by a general audience (think a business or research partner)
  • You can produce a report that is suitable for a professional audience (e.g., narrative is written in complete sentences, all graphs have proper titles and axis labels, there is not extraneous output, all Latex is rendered)
  • You can conduct the analysis using a reproducible data analysis workflow that incorporates version control

Review on logistic regression and Multinomial Logistic Regression

Analogize to a linear regression

  • Motivation
  • Model
  • Estimation
  • Interpretation
  • Inference
  • Comparison
  • Prediction
  • Condition
  • Issue
  • Extend to Multinomial logistic regression

Application exercise

Course Evaluation