Instructor:    Dr. Lingsong Zhang   lingsong@purdue.edu

Office hours Zoom Information: https://purdue-                                                        edu.zoom.us/j/99659684321?pwd=aCt4QzRSbzR6UnZnMHI0NTlpdEU3Zz09

Meeting ID: 996 5968 4321

Passcode: 730682

Office hours held by the instructor: Every Friday 10- 11 am

Teaching Assistant:  One TA will be assigned to you and help you throughout the semester. You may   contact your TA for R-related questions, website issues, exam settings, or homework-related questions. You may contact the instructor for methodology level and contextual questions or performance checks.  Email inquiries will be answered by the instructor and/or TAs within 24 hours or sooner in general during the weekday schedule but could be delayed during weekends and holidays.

The Designated TA and Contact Table

Office hour: there will be multiple sections of office hours each week. Check out the class website for more details.

Textbook (Required):  Special reprint of Applied Linear Statistical Models, 5th ed., by Kutner,    Neter, Nachtsheim &  Li, 2012 or 2013.  An Older version is also Okay and available from the library or book store.

Prerequisite: A course in general statistical methods that has calculus as a prerequisite (e.g., STAT    350, STAT 503, STAT 511). A self-check quiz is included in homework 1 to assess your understanding of some important concepts prior to taking this course. Please complete it on your own (open note) and determine if you are ready to take Stat512. There are some review materials on the course website.     Those who struggle with these concepts are strongly recommended to talk to the instructor for further   assistance.

Technical Requirements

The following information has been provided to assist you in preparing to use technology successfully.

•    A reliable internet connection - capable of consistently streaming video and stable enough to finish short exams without dropping connection.

•    Access to Purdue's Brightspace Learning Management System - All course content, course readings, and exams will be accessed online through Brightspace.

•    R and R-studio https://rstudio.com/products/rstudio/download/

•    The Respondus Lockdown browser is required to take the online exam. Check the appendix for more details.

Learning Objectives

This course focuses on the practical application of linear regression techniques rather than their      mathematical derivation. However, we use math extensively throughout the course, and most                     assignments will include a quantitative component.  Real world analyses usually require judgement  calls, and are very rarely amenable to rote applications of statistical techniques. Knowing when to use a  certain method, why it should (or should not) be used, and how to interpret its results are all at least as  important as knowing how to carry out the calculations to get those results or which numbers to look at  in the           output from a computer program. A solid conceptual understanding of the theory that underlies  linear        regression is critical to developing this knowledge. To do well in this class, you should aim to  learn not     only what to do in a given situation, but why.

After completing this course, you should be able to:

1.   Understand and explain the mathematical structure of linear regression models

2.   Build, analyze, and apply both simple and multiple regression models in real-world settings

3.   Conduct analysis of variance and covariance for regression

4.   Construct and interpret polynomial and interaction models

5.   Use linear regression models to perform statistical inference

6.   Identify the assumptions made by linear regression, use regression diagnostics to check those assumptions in the context of a particular application, and (if necessary) apply appropriate        remedial measures to reduce violations of the assumptions

7.   Conduct residual analysis

8.   Use data transformations to improve model performance

9.   Use model selection criteria to identify the best model from a set of candidate models

10. Use R to conduct regression analyses

Summary of expectations

I expect you to take responsibility for your own learning, to make an honest attempt to use the  resources available to you (lectures, the textbook, homework assignments, etc.) to master the course material, and to maintain the highest standards of academic integrity. I also expect you to attend class regularly, to come to each class (or listen to lecture recording) prepared to listen and answer               questions, and to behave appropriately  and politely toward other members of the class.

To prepare for class, it is often helpful to read  ahead of the lectures in order to get a general      introduction to the material. When doing this, try not to  get bogged down in details or equations.           Instead, focus on big picture concepts, becoming familiar with  terminology, and identifying any topics    that you find particularly challenging or difficult to understand. After lecture, while working on homework, or studying for exams, a second, more detailed reading of  specific sections of the text will help solidify   your understanding.

I will do my best to explain the material in a way that makes sense, but I will sometimes need      feedback  from you to know whether you understand a particular point or to figure out the best way to    explain a concept. If anything is unclear or confusing, or if you need help on a homework assignment, I expect  you to ask questions.  Similarly, the graders and I will do our best to evaluate your work in a fair, consistent, and transparent manner. However, mistakes do happen. If you are concerned that there has been an error in grading or if you do not understand why you received a particular score, I expect you to bring the issue to my attention in a timely manner.

Grading

Class on Track Credit   10%

Homework   30%

Exam 1   20%

Exam 2   20%

Independent case study

Total                                   100%

The final percentages needed for a particular grade are as follows:

98 – 100 = A+, 92-97=A, 90-91=A-

88 – 89 = B+, 82-87= B, 80-81=B-

78 – 79 = C+, 72-77= C, 70-71=C-

68 – 69 = D+, 62-67= D, 60-61=D-

0 – 59 = F.

The grades will not be curved.

Grading inquiries on most of the items must be addressed to your TA or instructor by Monday    8/1, except for the last two items: midterm 2 and the independent case study. After 8/1, the grade book will not be modified.

Unless there is a calculation error, final grades will not be changed. With that being said, “This is my last semester” , “I won’t graduate if I get so and so grade” , “ I will lose my scholarship” , “I am on probation” ,   or “I need a certain grade to maintain my financial assistance” ARE NOT LEGITIMATE REASONS TO

BUMP UP FINAL GRADES.

Class on track Credit

To earn the full credit each week, you need to watch the lecture recording no later than the end of the Sunday of the same week.

Homework: there are 6 Homework assignments, they are usually due on Tuesdays in Brightspace. They will be posted on Brightspace at least 1 week beforehand.

Homework must be submitted on Brightspace.  The format is required as below:

•    Copy the question, then work on it below.

•    Equations should be typeset using the Microsoft Word equation editor, MathType, or  LaTeX. If you are not familiar with any of these tools and need help getting started, please ask.

•    Include your name as it appears in Brightspace and the title of the  assignment at the top of the first page.

•    Answer questions in order, clearly label the answer for each question, and answer in complete English sentences unless otherwise specified in the question.

•    Provide any necessary graphs or analytical results (and only the necessary results) with  the    relevant question. Please do not place them at the end of the assignment. Every graph or        table should be accompanied by a title or caption containing a label (e.g., “Fig.  1” or “Table 1”) and briefly describes the contents. See the example on Brightspace for details.

•    Paste only relevant output and explain major findings from the output. An unexplained output will be considered incomplete and 50% will be deducted from the problem.

•     Provide sufficient working details for every section and subsection of the question. The graders are not supposed to trace the details all around the place. Graders reserve the right of final explanation

•    A copy of your R code must be appended to each homework question and labeled clearly.  In the event that there is an error, this will help the grader to identify the problem and assign      appropriate partial credit.  If the code is not included,  50% will be deducted from the               assignment.

•    If you choose to work on the entire homework via R markdown, please knit it into a word document and submit it.

•    If multiple submissions are identical or highly similar, all will receive no grade. Graders reserve the right of the final explanation

•    If a submission cannot be opened in Brightspace or a wrong version is submitted, it will receive no grade.

•    No late homework will be accepted.

Self-check quiz: The self-check quiz serves two purposes:

•    Evaluate your current skill set (the prerequisites) on hypothesis test components and some

fundamental theories on sampling distribution and population distribution.  If you struggle with   some of the concepts and fail to compete within one hour, you may review some of the relevant material under week 1; or discuss it with the instructor.

•    Set up the exam environment properly and make sure you are familiar with the exam procedure.

Specifically, you will need to download the Respondus lock down browser and also need a webcam and microphone.

The Respondus lock down browser can be downloaded through the following link:

https://www.purdue.edu/innovativelearning/supporting-instruction/instructional-technology/respondus- browser.aspx

Some instruction of the Respondus lock down browser and exam is attached at the end of the syllabus.

The self-check quiz has to be completed by the end of Friday, 6/16. Once start, you will have one hour to complete. This quiz is graded by completion.

Exam: There will be two online exams during the semester.  Check the course website for the detailed schedule

All questions are multiple-choice type that covers essential concepts in the topic. The exams   focus on conceptual understanding and application of essential methods, and are not computationally comprehensive.  Concepts are well covered in the assignments and examples available on the course website.  No sample exam will be available, but some questions could be demonstrated in class.

One A-4 or letter-paper size paper can be used as a cheat sheet for each of the exam. The cheat sheet must be hand-written. Anything else (scratch paper, writing tool of any kind, calculator) is not         allowed.

You must independently complete the exam, and should not discuss it with anyone until your score is  available in Brightspace.  

The exam process will be recorded by the Respondus monitor and later checked by the instructor or TAs. Absence without permission of the instructor will result in no grade and no make-up exam will be available.

Independent Case Analysis:  In order to access the practical skill of data analysis with linear regression methods, a final independent case study is scheduled for the last week.

You must complete the case study independently or receive no grade. The instructor and the TA reserve the right to the final decision.

The topic of the case study will be emailed to you by your designated TA shortly before Monday 4pm, and due on 4pm on Tuesday.

The Case study has a similar format requirement as homework.

Any question regarding to the exam other than the administrative level will not be answered during the exam period. The instructor and the TA reserve the right for the final decision.

Academic integrity and collaboration

PURDUE HONORS PLEDGE

As a boilermaker pursuing academic excellence, I pledge to be honest and true in all that I do.

Accountable together - we are Purdue.

I strongly encourage collaboration that genuinely aims to improve your conceptual understanding and retention of the course material.  This includes group study sessions, discussing homework                 assignments, and comparing your answers with other students.  However, copying answers or            computer code from another student or from a similar problem assigned in a previous semester does  not serve any learning objective, and will be penalized as academic misconduct. Any attempt to cheat on an exam also will be treated as academic misconduct.

Any student who is found to have engaged in academic misconduct on any assignment, project, or exam, or to have enabled misconduct by another student, will receive penalties up to and       including an F for the course and will be referred to the Dean of Students for potential further     sanctions.  There will be no exceptions.  Please note that “enabling misconduct” includes                  unauthorized attempts to make assignments, exams, or answer keys available to students in future       semesters.

If you are unsure whether a particular action constitutes misconduct, please feel free to ask. A good rule of thumb is to talk with other students about how to do a problem, do the work and write it up by yourself, and then compare answers afterwards, repeating the cycle as needed to correct any errors.

Appendix I. Instruction for Respondus Lockdown Browser and exam procedure

Respondus LockDown Browser is an exam proctor system. You will take the quizzes through Respondus LockDown Browser this semester in Stat512.

Please follow the simple guide to install the Respondus Lockdown browser.

If you run into any issues, contact TAs for help as soon as possible.

Install

You may use the link below to download Respondus LockDown Browser

https://www.purdue.edu/innovativelearning/supporting-instruction/instructional-technology/respondus- browser.aspx

1    Click DOWNLOAD RESPONDUS BROWSER (INCLUDES MONITOR)

2    Then, choose INSTALL NOW

3    Run the download file to install Respondus LockDown Browser

Start exam

1    Using your original browser (NOT Respondus LockDown Browser), go to Brightspace, log in and enter course webpage.

2    Click Course Tools, go to Quizzes and choose the exam you are going to take.

3    Click Launch LockDown Browser at the bottom

4    Follow the instruction in LockDown Browser