MATH 4387/5387 001 Applied Regression Department of Mathematical and Statistical Sciences University of Colorado Denver College of Liberal Arts and Sciences Spring 2016 Professor: Joshua French Office: AB1-4213 Phone: 303-315-1709 Email: joshua.french@ucdenver.edu Course website: http://canvas.cuonline.edu Twitter: https://twitter.com/dr_jfrench (this usually has statistics/r related tips, information I ve found helpful). Class Meeting Times: Tuesday and Thursday from 12:30-1:45PM in AB1-4125. Office Hours: Tuesday and Thursday from 2:00-3:00 and by appointment. Course Description: Topics include simple and multiple linear regression, model diagnostics and remediation, and model selection. Emphasis is on practical aspects and applications of linear models to the analysis of data in business, engineering and behavioral, biological and physical sciences. Overview: Regression analysis is the most commonly used statistical analysis tool. In this course you will learn how to fit a regression model to data and then make statistical inference about the model. The statistical inference is based on various assumptions, so we will learn how to assess model fit and determine whether the appropriate assumptions are met. Often, more than one model will appear to fit the data reasonably well, so we will also learn various approaches to selecting a good model. Lastly, we will learn some basic aspects of experimental design and analysis. The course will involve using the R statistical software package to do much of the analysis. Some proofs will be required for the course, and examples of the expected types of proofs will be supplied during the course. Course Goals and Outcomes: Overall Learning Objectives Understand and perform simple and multiple linear regression modeling, including the additional of complex regressors in the model. Learn how to perform statistical inference for both parameters and responses in regression models Learn how to assess whether a regression model satisfies the applicable assumptions Learn how to choose a good model and assess validity Joshua French Page 1 of 9 Syllabus
Learn how to deal with problems commonly faced in regression related to predictors, errors. Learning Outcomes Several specific learning outcomes will be strengthened in this course. Among them are: Critical Thinking Inquiry and Analysis Quantitative Literacy Critical thinking will be necessary to some degree in every course assignment. However, this will be especially true in homework assignments after Exam 1, as well as Exam 2 and the final exam. Specifically, when making statistical inference, one will have to weigh the evidence before accepting a conclusion. Inquiry and Analysis will be required in every assignment. All of these assignments will require you to determine the question of interest and evaluate evidence provided by the data before coming to a conclusion. Quantitative Literacy will be promoted in all homework assignments, exams, and some quizzes. Each of these assignments will require you to work with data and numbers to answer questions and convincingly argue for certain conclusions. Major Topics Regression analysis is the most commonly used statistical analysis tool. In this course you will learn how to fit a regression model to data and then make statistical inference about the model. The statistical inference is based on various assumptions, so we will learn how to assess model fit and determine whether the appropriate assumptions are met. Often, more than one model will appear to fit the data reasonably well, so we will also learn various approaches to selecting a good model. Lastly, we will learn some basic aspects of experimental design and analysis. The course will involve using the R statistical software package to do much of the analysis. Some proofs will be required for the course, and examples of the expected types of proofs will be supplied during the course. Rationale Data analysis skills are extremely important in today s workforce. And linear regression is one of the most basic theoretical model building foundations used in more complex methods. You will learn the fundamentals of linear regression, data analysis, and statistical programming in this course. In contrast to introductory courses that sometimes focus on mechanistic application of statistical tools, this course will help you to think like a data analyst, understand the foundations of linear regression, and provide you with the tools to continue growth in statistical training. This course should be useful for future courses and in the work force. Joshua French Page 2 of 9 Syllabus
Prerequisites: MATH 3191 and MATH 3800 or 4820. Some computer programming experience. Required Textbook: Linear Models with R, 2 nd edition by Julian Faraway. ISBN: 978-1439887332 Optional Textbook: Applied Linear Regression, 4 th edition by Sanford Weisberg. ISBN: 978-1-118-38608-8. Computing: We will use the freely available R statistical software for data analysis in this class. R is a free, cross-platform statistical package that is extremely powerful. It is the standard statistical computing language in academia and becoming more popular in the public and private sector. The language is somewhat similar to Matlab. R may be downloaded at www.r-project.org. Even better, download Microsoft R Open, which optimizes the linear algebra library used for computations. We will cover all of the necessary details in class. RStudio is a free and open source integrated development environment (IDE) for R. It is also cross-platform and has many nice features (like the ability to scroll through your plots, color coding, see the variables in your environment, etc.) that you may find useful. It may be downloaded at www.rstudio.org. Grades: The course is cross-listed as both an undergraduate and graduate course. Accordingly, the grading policies are slightly different for each course as specified below. Homework: Homework problems will be assigned for each section of material. On the designated days, students are expected to turn a physical paper copy of their completed homework into the instructor by the end of class. Some of the assigned problems will be selected for grading (as few as one problem and as many as all of them). These problems alone will be used to determine a student s grade for that assignment. Exams: There will be two exams during the semester. The first exam will take place on Thursday, February 25 th and the second exam will take place on Thursday, April 14 th. Exam due dates are fixed now so plan accordingly. Without prior approval or a documented health, disability, or emergency reason, missed quizzes will be scored as a zero. Final Project: All students are expected to complete a final project. Undergraduate students will work in groups of a designated size (to be determined later in the semester) to complete the project and graduate students will complete their work individually. Students will analyze a data set of interest, describe their results in a formal paper, and present their results to the class in a short presentation during the time scheduled for the final exam. The project grade will be determined on the basis of the accuracy of the statistical analysis and the quality of the paper and presentation. More details about this project will be given at a later time. MATH 5387: Students enrolled in MATH 5387 may be assigned extra homework and exam problems. These problems are not optional for students enrolled in MATH 5387. When explicitly stated, MATH 4387 students may do these problems for extra credit. Joshua French Page 3 of 9 Syllabus
Final grades will be determined according to the following weighting scheme: Homework 20% Final Project/Presentation 20% Exam 1 30% Exam 2 30% Letter grades will be determined by the following scale: Percentage Letter Grade Percentage Letter Grade 92 or higher A 78 up to 80 C+ 90 up to 92 A- 70 up to 78 C 88 up to 90 B+ 68 up to 70 D+ 82 up to 88 B 62 up to 68 D 80 up to 82 B- 60 up to 62 D- Below 60 F Grades will be provided via the course s Canvas course shell. You can access your scores at any time within the Canvas gradebook. Homework, quizzes, and exams will be returned in class or in my office directly to you. Academic Honesty Policy: Students are responsible for completing all assignments without assistance (either voluntary or involuntary) from other students unless otherwise directed by the instructor. The minimum punishment for any form of cheating is a zero on the assignment, quiz, exam, etc. If you have a question regarding the wording of a problem on a quiz or exam, you may ask me to assist you with the wording. Discussion of the homework assignments between students can be helpful, and is encouraged. Homework assignments should not be copied and work should not be shared. Plagiarism is the use of another person s words or ideas without crediting that person. Plagiarism and cheating will not be tolerated and may lead to failure on an assignment, in the class, and dismissal from the University. For more information see the Academic Honesty Handbook. Students with Disabilities: The University of Colorado Denver is committed to providing reasonable accommodation and access to programs and services to persons with disabilities. Students with disabilities who want academic accommodations must register with Disability Resources and Services (DRS), North Classroom 2514, Phone 303-556- 3450, TTY 303-556-4766, Fax 303-556-4771. DRS requires students to provide current and adequate documentation of their disabilities. Once a student has registered with DRS, DRS will review the documentation and assess the student s request for academic accommodations in light of the documentation provided. Once you provide me with a copy of DRS s letter, I will be happy to provide the accommodations DRS has approved. Student Code of Conduct: All students are expected to abide by the University of Colorado Denver Student Code of Conduct. Joshua French Page 4 of 9 Syllabus
Absences, Tardiness, Homework, Quizzes, and Exams: Except for documented health, disability, or emergency reason or prior approval, I will not accept excuses for absences or tardiness. Unless otherwise specified, missed or late homework, quizzes, or exams will be scored as a zero. Documentation of disability or health related issues must be provided to Disability Resources and Services, North Classroom 2514, Phone 303-556-3450, TTY 303-556-4766, Fax 303-556-4771. Drops and incompletes: You have until October 28th to drop the course with only the instructor's (but not a Dean's) signature. The incomplete policy of the department and college is strictly enforced. Incompletes are given only in situations in which a student has: (1) Successfully completed 75 percent of the course (i.e. is passing the course) (2) Special circumstances (verification may be required) that preclude the student from attending class and completing graded assignments, and (3) Made arrangements to complete missing assignments with the original instructor. A CLAS Course Completion agreement is strongly suggested. Incompletes are not granted for low academic performance. Expectations and Other Details: This is a senior level/graduate level course. Expect to work hard. Expect to work between 6 to 9 hours per homework assignment. Expect to study for tests and quizzes. Class attendance and participation is expected. Students are expected to enter class on time and remain in class for the duration of the class period. Students are expected to ask questions (during class or office hours) if they are confused. It is your responsibility to ask for help! Students are expected to read the relevant sections of the book. The book will go in greater detail than I am able to go in class and will be helpful in understanding the lecture material. Material should be read BEFORE the appropriate lectures. Homework should be completed as the homework sets are assigned. Students will find little benefit in rushing to complete homework assignments. Course announcements will be made frequently through Canvas and/or email. Emails will be sent to your ucdenver.edu email address in accordance with university policy. You are responsible for the information contained in any announcements or messages I send you, regardless of whether the information is repeated in class. It is your responsibility to frequently check Canvas and to maintain your university email address. Turn off or silence cell phones during class. If you have a conflict between class and a major religious observance, please talk to me in advance so that we can make the appropriate accommodations. Student Grievances: Students with complaints about the course or instructor should: 1) meet with the instructor face-to-face; 2) if not satisfied, meet with Steve Billups (Associate Chair, primary) or Jan Mandel (Chair); 3) if not satisfied, appeal to the Associate Dean. No step in this process may be skipped. See "Procedures for Student Grievances about Courses or Faculty, CLAS" Joshua French Page 5 of 9 Syllabus
University wide policies: Student Code of Conduct http://www.ucdenver.edu/life/services/standards/students/pages/default.aspx Accommodations http://www.ucdenver.edu/student-services/resources/disability-resourcesservices/accommodations/pages/accommodations.aspx Academic Freedom http://www.ucdenver.edu/policy/pages/academic-freedom.aspx Family Educational Rights and Privacy Act (FERPA) http://www.ucdenver.edu/studentservices/resources/registrar/students/policies/pages/studentprivacy.aspx Attendance http://www.ucdenver.edu/faculty_staff/employees/policies/policies%20library/oaa /StudentAttendance.pdf Discrimination and Harassment Policy and Procedures http://www.ucdenver.edu/about/whoweare/chancellor/vicechancellors/provost/st udentaffairs/universitylife/sexualmisconduct/denverpolices/pages/denverwelcome. aspx Grade Appeal Policy http://www.ucdenver.edu/policy/documents/process-for-grade-issues.pdf I reserve the right to modify this syllabus as the semester progresses. Joshua French Page 6 of 9 Syllabus
Tentative Course Schedule: The following course schedule is tentative. The exam due dates are fixed but the material covered on the exams is subject to change. Homework assignment due dates will be determined by how quickly we move through material and will be given throughout the semester. I reserve the right to modify this schedule as the semester progresses. Date Day Week Tentative Agenda Topic/Reading Assignment Due 19-Jan T 1 Syllabus/Introduction Ch 1 21-Jan R 1 Introduction Ch 1 26-Jan T 2 Estimation Ch 2 28-Jan R 2 Estimation Ch 2 Hw 1 2-Feb T 3 Estimation Ch 2 4-Feb R 3 Explanation/Interpretation Ch 5 Hw 2 9-Feb T 4 Explanation/Interpretation Ch 5 11-Feb R 4 Random Variables Ch 0 Hw 3 16-Feb T 5 Inference Ch 3 18-Feb R 5 Inference Ch 3 Hw 4 23-Feb T 6 Inference Ch 3 25-Feb R 6 Exam 1 1-Mar T 7 Prediction Ch 4 3-Mar R 7 Prediction Ch 4 8-Mar T 8 Complex Regressors Ch 5/ALR 10-Mar R 8 Complex Regressors CH 5/ALR Hw 5 15-Mar T 9 Diagnostics Ch 6 17-Mar R 9 Diagnostics Ch 6 Hw 6 22-Mar T Spring Break 24-Mar R Spring Break 29-Mar T 10 Diagnostics Ch 6 31-Mar R 10 Diagnostics Ch 6 Hw 7 5-Apr T 11 Model Selection Ch 10 7-Apr R 11 Model Selection Ch 10 Hw 8 12-Apr T 12 Problems with Predictors Ch 7 14-Apr R 12 Exam 2 19-Apr T 13 Problems with Predictors Ch 7 21-Apr R 13 Problems with Errors Ch 8 26-Apr T 14 Problems with Errors Ch 8 Hw 9 28-Apr R 14 Transformation Ch 9 3-May T 15 Project preparation Hw 10 5-May R 15 Project preparation May 9-14 Project presentation Joshua French Page 7 of 9 Syllabus
Spring 2016 CLAS Academic Policies The following policies, procedures, and deadlines pertain to all students taking classes in the College of Liberal Arts and Sciences (CLAS). They are aligned with the Official University Academic Calendar: http://www.ucdenver.edu/student-services/resources/registrar-dev/courselistings/pages/academiccalendar.aspx Schedule verification: It is each student s responsibility to verify that their official registration and schedule of classes is correct in their Passport ID portal before classes begin and by the university census date. Failure to verify schedule accuracy is not sufficient reason to justify late adds or drops. Access to a course through Canvas is not evidence of official enrollment. E-mail: Students must activate and regularly check their official CU Denver e-mail account for university related messages. Administrative Drops: Students may be administratively dropped from a class if they never attended or stopped attending, if the course syllabus indicates that the instructor will do this. Students may be administratively dropped if they do not meet the requisites for the course as detailed in course descriptions. Late adds and late withdrawals require a written petition, verifiable documentation, and dean s approval. CLAS undergraduate students should visit the CLAS Advising Office (NC1030) and graduate students should visit the Graduate School (12 th floor LSC) to learn more about the petition process and what they need to do to qualify for dean s approval. Waitlists: The Office of the Registrar notifies students at their CU Denver e-mail account if they are added to a class from a waitlist. Students are not automatically dropped from a class if they never attended, stopped attending, or do not make tuition payments. After waitlists are purged, students must follow late add procedures to be enrolled in a course. Students will have access to Canvas when they are on a waitlist, but this does not mean that a student is enrolled or guaranteed a seat in the course. Students must obtain instructor permission to override a waitlist and this is only possible when there is physical space available in a classroom, according to fire code. Joshua French Page 8 of 9 Syllabus
Important Dates and Deadlines All dates and deadlines are in Mountain Time (MT). January 19, 2016: First day of classes. January 24, 2016: Last day to add or waitlist a class using the Passport ID portal. January 24, 2016: Last day to drop a class without a $100 drop charge--this includes section changes. January 25, 2016: All waitlists are purged. Students should check their schedules in their Passport ID portal to confirm in which classes you are officially enrolled. January 26-Feburary 3, 2016, 5 PM: To add a course students must obtain instructor permission using the Instructor Permission to Enroll Form and bring it to the CLAS Advising Office (NC 1030) or have their instructor e-mail it to CLAS_Advising@ucdenver.edu. February 3, 2016: Census date. o 2/3/16, 5 PM: Last day to add full term classes with instructor approval. Adding a class after this date (late add) requires a written petition, verifiable documentation, and dean s approval. After this date, students will be charged the full tuition amount for additional classes added College Opportunity Fund hours will not be deducted from eligible student s lifetime hours. o 2/3/16, 5 PM: Last day to drop full term classes with a financial adjustment on the Passport ID portal. After this date, withdrawing from classes requires instructor signature approval and will appear on student s transcript with a grade of W. After this date, a complete withdrawal (dropping all classes) from the term will require the signature of the dean and no tuition adjustment will be made. Students should consult appropriate service offices (e.g. international status, Financial Aid (loans, grants, and/or scholarships) or Veteran s Student Services) before withdrawing from course(s) to determine any impact for continued enrollment and funding. o 2/3/16, 5 PM: Last day to apply for Spring 2016 graduation. Undergraduates must make an appointment and see their academic advisor before this date to apply. Graduate students must complete the Intent to Graduate and Candidate for Degree forms. o 2/3/16, 5 PM: Last day to request No Credit or Pass/Fail grade for a class using a schedule adjustment form. o 2/3/16, 5 PM: Last day to petition for a reduction in Ph.D. dissertation hours. February 4-April 4, 2016, 5 PM: To withdraw from a course, students must obtain instructor permission using the Schedule Adjustment Form and must bring the signed form to the Office of the Registrar. To add a course, students must petition through College/School undergraduate advising offices or the Graduate School, as appropriate. March 21-27, 2016: Spring break- no classes, campus open. April 5, 2016: The Office of the Registrar now requires both the instructor s signature and a CLAS advisor s/dean s signature on a Schedule Adjustment Form to withdraw from a class. Students should consult their home college advising office for details. April 18, 5 PM: Deadline for undergraduate CLAS students to withdraw from a course without filing a late withdrawal petition. Contact CLAS Advising (NC 1030 303-556-2555). May 14, 2016: End of semester. June 24, 2016: Final grades available on the Passport ID portal and on transcripts (tentative). Please contact an academic advisor if you have questions or concerns. Joshua French Page 9 of 9 Syllabus