NEW COURSE PROPOSAL USC Columbia, Lancaster, Salkehatchie, Sumter & Union campuses NCP INSTRUCTIONS: This form is used to add a new course to the University course database. This form is available online at www.sc.edu/provost/acadprog. Date: 10/22/2014 Campus: columbia College/School: Arnold School of Public Health Department (if applicable): Epidemiology & Biostatistics Undergraduate Graduate PROPOSED COURSE INFORMATION Course Designation: Bios 755 4-letter Designator Prefix Course Number Suffix # Credit Hours: 3 Variable Fixed # Times Course Can Be Taken: 1 Course Title (24 character limit): Introduction to Longitudinal Data Analysis Cross-listed with which course? 4-letter Designator Prefix Course Number Suffix Course Description: (50-word limit) Introduction to principles and methods for longitudinal data. Focus on data analysis and interpretation. Course Prerequisites/Corequisites: Bios 757 Course Delivery Location: USC Campus Off-Campus site (If off-campus delivery is being requested, attach a completed Off-Campus Delivery (OCD) form.) Course Delivery Method: Traditional Delivery Distance Technology Delivery (streaming video, web-based, CD/DVD) (If distance technology delivery is being requested for the first time, attach a completed Distance Education Delivery (DED) form.) Proposed Effective Term - Change to database/bulletin effective no sooner than: Year: 2015 Fall Spring May Session Summer I Summer II Required Resources: Does this course require additional faculty, facilities, library resources or funding? Yes No (If yes, attach letters of commitment from appropriate official(s).) Grading System: Standard Pass/Fail Only Not Auditable Rationale for grading system other than standard: Enrollment Restrictions: Restricted to: Excluded: Special Permissions required? Department Instructor
JUSTIFICATION FORM USC Columbia, Lancaster, Salkehatchie, Sumter & Union campuses JUS INSTRUCTIONS: Please attach a statement explaining the justification for the proposed program or course action. This form is available online at www.sc.edu/provost/acadprog. Date: 10/22/2014 Campus: Columbia College/School: Arnold School of Public Health Department (if applicable): Epidemiology & Biostatistics Degree Program (if applicable): Undergraduate Graduate The department of Biostatistics has had numerous inquires for Bios 770 (longitudinal Analysis) from students outside of our department. Many of these students were advised not to take Bios 770 because their background was insufficient. We are proposing Bios 755 as a course covering the basics of Bios 770, but at a non-biostatistical masters level. This way students from other departments could benefit. University of South Carolina / Provost Office / Justification Form Oct 2010 v3
Suggested listing BIOS 755 Introduction to Longitudinal Analysis [Print Course] BIOS 755 Introduction to Credits: 3 introduction to principles and methods for longitudinal data, which are often encountered in practice where multiple measures are observed over time on an individual. This course is designed for non-biostatistics major researchers, with a focus on data analysis and interpretation more than theoretical development. Problems will be motivated by applications in epidemiology and clinical medicine, health services research, and disease natural history studies. This course is for non-biostatistics majors. Prerequisites: BIOS 757
BIOS 755 Introduction to Longitudinal Data Analysis Spring 2016 Time: TBA Instructor: TBA Office: TBA e-mail: TBA Office Hours: TBA Required Textbook: Applied Longitudinal Analysis, by Fitzmaurice, Laird & Ware Suggested Readings: Analysis of Longitudinal Data, by Diggle, Heagerty, Liang and Zeger, 2 nd ed. Course Description: This course provides an introduction to principles and methods for longitudinal data, which are often encountered in practice where multiple measures are observed over time on an individual. This course is designed for non-biostatistics major researchers, with a focus on data analysis and interpretation more than theoretical development. Problems will be motivated by applications in epidemiology and clinical medicine, health services research, and disease natural history studies. This course is for non-biostatistics majors. Course Learning Outcomes: A student who successfully completes this course will be able to- 1. Understand the need for longitudinal and hierarchical data analysis and why analytic techniques that ignore the longitudinal nature of the data are inappropriate. 2. Understand, specify and apply longitudinal and hierarchical models as a means of addressing scientific and public health issues. 3. Identify which statistical models and methods to apply for the analysis of longitudinal data. 4. Apply and interpret the results of statistical analyses for longitudinal data using SAS and R. 5. Successfully collaborate with statisticians regarding more complex data analyses, and critically appraise statistical procedures used in selected research articles. Grading: Homework: 55% Midterm Exam: 20% Final Exam: 25% Grades will be assigned as follows: A = 90-100, B = 80-89, C = 70-79, etc. A + may also be amended if so warranted. You are allowed to work together on homework assignments, as long as your homework does not become simply a copy of another. However, exams are to be entirely your own work. Homework will be assigned frequently. Assignments will generally be due one week from the date assigned, and will usually be graded and returned during the class that follows. Assignments will be due at the beginning of class on the date due (regardless if you will be attending that class). Assignments turned in after homework is collected before the return date, will have the grade reduced by 50%. Assignments will generally not be accepted after the return date. 1
The lowest homework grade will be dropped. The final exam will be given on XX XX from XX- XX. Cheating: No form of cheating or plagiarism will be tolerated in any way. All students are expected to read, understand, and abide by the Carolinian Creed, which is given on the back. SAS: If you wish to purchase an annual license of SAS for your home computer, it is available through VIP (or my.sc.edu). Disabilities: Reasonable accommodations are available for students with a documented disability. If you have a disability and may need accommodations to fully participate in this class, contact the Office of Student Disability Services in LeConte College Room 112A: 777-6142, TDD 777-6744, email: sasds@mailbox.sc.edu, website: http://www.sa.sc.edu/sds. All accommodations must be approved through the Office of Student Disability Services. 2
Topics (# of classes) Introduction and Examples (1) Basic Covariance Matrix and ANOVA (1) Covariance Structure (2) Models for the Mean (2) General Linear Model for Longitudinal Data (1) Maximum Likelihood Method and Restricted Maximum Likelihood Method (1) Random Coefficient Models (2) Linear Mixed Effects Models (2) Generalized Linear Models (1) Generalized Linear Mixed Models (1) Logistic Mixed Effect Models (1.5) and Poisson Mixed Models (0.5) Marginal Models (1) Generalized Estimating Equations (GEE) (1) Multilevel Models (3) Variable Selection and Model Selection (1) Missing Data in Longitudinal Data Analysis (2) Sample Size Estimation in Longitudinal Data Analysis (1) Total classes: 25 Optional topics: Study Design Analysis of Experimental Studies Nonlinear Mixed Effects Models Semiparametric Models with Longitudinal Data 3
Recommended Text: Fitzmaurice, Laird & Ware. Applied Longitudinal Analysis, 2nd edition. SAS Press Series. ISBN: 978-0470380277 Optional Reference texts available online for downloading: SAS Online Doc - http://support.sas.com/documentation/cdl_main/index.html Optional Software: SAS 9.2 (12-month license) for $60 each from USC s Computer Services Division. Call 777-4657 or visit http://www.sc.edu/software for more information.