Download Biostatistics Course: Learning Statistical Techniques for Public Health and Medicine and more Exercises Mathematics in PDF only on Docsity!
School of Public Health
PBHL 520
BIOSTATISTICS
4 hours credit
Basic Information
Course director: Ed Gracely, PhD Office: 6 th^ floor, Bellet Building - Room 611.
Office Phone: 215-991-8466 (at my main office in Queen Lane) Email: Egracely@drexelmed.edu. Good way to reach me. I check e-mail from everywhere. Office Hours: By Appointment Only
Instructor Ed Gracely PhD
Quarter Offered: Fall
Course Time: Tuesdays 2-3:50, Thursdays 2-3: See calendar.
Location: Lecture Hall A on Tuesday, Geary B on Thurs.
COURSE DESCRIPTION
This course introduces the basic ideas of biostatistics as used in public health research and practice. Biostatistics largely consists of the statistical methods used in any data analytic setting. It includes, for example, techniques for describing and summarizing observations, for assessing associations among variables, and for determining the extent to which chance may be explaining and/or influencing the observed results.
The material in this course is relevant to all aspects of public health, but it is especially closely allied with epidemiology, the science concerned with the occurrence, distribution, and causality of diseases and other health-related conditions in the world. Most people are familiar with epidemiologic research results, such as those on diet and cancer, or fitness and heart disease. All epidemiology investigations include biostatistics as a core component. Epidemiology will be the focus of a course in the next quarter.
Biostatistics is a required course for the Master of Public Health (MPH) Program of Study. Non public health students are welcome as well.
COURSE OBJECTIVES
Upon completion of this course, students should be able to:
1. Read and interpret the public health and medical literature involving data presentation and analysis
including:
1.()a Simple, descriptive analyses like means and other summary statistics, scatter plots, tabular
presentations
1.()b Techniques to quantify the margin for error, notably the SEM and confidence intervals
1.()c Procedures for estimating sample size and statistical power
1.()d Basic correlation and regression techniques
1.()e Survival curves
1.()f Common inferential techniques such as t-tests, nonparametric tests, and chi-square
1.()g Logistic regression and other multivariate procedures
2. Make appropriate choices among statistical techniques.
3. Perform some important techniques by hand, notably:
3.()a descriptive methods, including z-score calculations
3.()b t-tests, chi-square, and some basic nonparametric techniques
3.()c manipulations of the SEM and some basic confidence intervals
3.()d illustrative power/sample size determinations
3.()e correlation and regression calculations.
3.()f Simple life table/survival curve analysis.
4. Create a STATA dataset, save it, and perform simple analyses and reports based on it.
5. Conduct basic analytical functions in Excel.
6. Interpret and write clearly about the results of statistical analyses they perform, whether by hand or by
computer.
COURSE CONTENT
Introduction to Biostatistics is built around three types of scheduled activities, which are:
- Didactic sessions: these are the core of the class, in which material is taught by an expert. Most involve biostatistics, but several sessions are devoted to computer techniques in STATA, and one in Excel.
- Labs , to be held at selected times in the calendar. They mainly involve working together to solve practice exercises. They are held in the regular classroom. You will do the lab activities in class, NOT in advance. They are not collected or graded.
- Exams : See below.
There are also exercises for grade, which are not an "activity" but are an important part of the class.
All sessions are mandatory.
Handouts are posted on Blackboard. https://learn.dcollege.net/
Exams Students will complete a mid-term exam and a final exam. Study guides for each will be posted.
DT Ch 5 pp 121-122, 124 (Wilcoxon section on page 124). Ch 6 pp 143-146. Ch 8 pp 198- (correlation) Chi-square : DT Ch 6 pp 149- Survival DT Ch 9 pp 221- Logistic Regression DT Ch 10, pp 245-248 (general multiple variable ideas), 261-263 (LR itself).
Although there is no required text, there is required software:
For this class, you will need to purchase STATA. My official source sent this note:
STATA is available for purchase at GradPlan pricing (student pricing) by all students, staff, and faculty of Drexel University. The prices represent a substantial discount off the already-discounted academic price of Stata. The software offered through the Stata GradPlan is not student software; that is, it is in no way limited in terms of features.
Prices are (sales tax, if applicable, and shipping not included):
STUDENTS ONLY
Small Stata software with PDF documentation 35.00/six months
Small Stata software with PDF documentation 49.00/one year
Stata/IC software with PDF documentation 69.00/six months
How to order
Online: http://www.stata.com/order/new/edu/gradplans/campus-gradplan Email: bkrauss@stata.com Phone: 800-782-8272 (Monday-Friday, 8 to 5 Central Time) Fax: 979-696-
Delivery is available via electronic software download or DVD. Be sure to include your DREXEL.EDU email address when ordering.
Method of payment
o Visa, MasterCard, American Express, Discover, JCB, or Diner's Club o Purchase order o Check payable to StataCorp with order
Small STATA may be enough for class, but it is limited to 1,200 observations. So 12 variables on 100 subjects meets its limits, and it may not be adequate for any actual data analysis.
Note that STATA is not part of Drexel's software collection. A limited number of copies will be available in the SPH Smart Room on the third floor of Bellet, but most students should plan on purchasing a copy.
TEACHING METHODS
This course is taught through a combination of didactic lectures and labs. Assigned text readings will broaden the students’ understanding of topics covered in class. Each student will do a number of review exercises. Students will complete two in-class exams that cover all of the material presented during this course.
Regular attendance and active participation in class discussions are essential for achieving a meaningful learning experience. Showing up on time is a matter of professionalism. It is disruptive to other students to come after class has started. For exams, being in class by the start time is essential.
EVALUATION METHODS
The final grade will be based on exams, general exercises, and Computer exercises (STATA and Excel). Distribution of the final grade is as follows:
EVALUATION METHOD
PROPORTION OF FINAL
GRADE
STATA and Excel Exercises 10% General Exercises 15% Exams 75%
Assignments: Computer assignments and exercises will be posted on Blackboard well before they are due. Due dates are indicated on the instructions and on the course calendar. Exercises should be brought to class on the due date. Please staple them if they are > 1 page! Put your name on every page too. One Excel exercise is to be e-mailed, however. Please keep each set of exercises separate even if they are due on the same day. I may ask one TA to grade one set, and a different TA to grade the other. Exercises must be handed in on time, or appropriately excused. You may get a 0 for an exercise if it is handed in late without a good reason. You are expected to be pro-active about this. Tell me in class, or by e-mail, a.s.a.p. if you realize you won't have the exercise on time. I might even accept, "I left it home, can I bring it tomorrow", once. But if I have to e-mail you to find out why I don't have it, the explanation will have to be much stronger. You may work with other students on all exercises. The product you hand in must, however, be your own work. In other words, your contact with others should be verifying each others’ independent work, helping someone who can’t do it to understand the problem and solve it, etc… Do NOT simply copy each others’ results. This is true for all exercises, including computer exercises. The work you hand in must be your own.
Grading : The final grade is A, B, C or F based on the combined total points from exercises and exams. 70% will be the minimum for attaining a grade of C, 80% for a B, and 90% for an A-. +/- grades will be used at the instructor’s discretion. There is no F+ grade! (And no D’s). I don't round at all. 89.9 is a B+, 90 is an A-. There are a lot of A's in this class, and I need to draw the line somewhere. I do use C-, which is better
Academic Honesty Policy: Drexel University is committed to a learning environment that embraces academic honesty. In order to protect members of our community from results of dishonest conduct, the University has adopted policies to deal with cases of academic dishonesty. Please read, understand, and follow the “Academic Honesty Policy” as written in the Official Student Handbook.
http://drexel.edu/studentaffairs/community_standards/studentHandbook/general_information/ code_of_conduct/
PLAGIARISM STATEMENT Students will not plagiarize or otherwise use or present the work of others (published, copyrighted, or other material) as their own work without appropriate references and/or attribution. This applies to all activities, including development and presentation of ‘Learning Issues’.
Adopted by the Educational Coordinating Committee
July 26, 2000
Course calendar
LHA - Tuesdays Geary B - Thursdays Sept 24 Week 1 Didactic: Introduction Didactic: Descriptive Statistics
Sept 26 Didactic : Descriptive Statistics part 2 Excel lesson (Intro, simple examples) Oct 1 Week 2 Didactic: Distributions Didactic : Accuracy of a mean/SEM
Oct 3
Didactic : Graphs & Tables
STATA (Intro to STATA programming) Oct 8 Week 3 Didactic Hypothesis Testing
Oct 10 Lab: Exercise due: Descriptives/Distributions/SEM Oct 15 Week 4 Didactic t-tests
Oct 17 Didactic Chi-square Exercise due: Graphs and Tables Exercise due: Computer exercise 1 (Excel) Oct 22 Week 5 Didactic Confidence intervals
Oct 24 Didactic : Nonparametric tests 10- Exercise due: Hypothesis tests and t-tests Exercise due: Computer exercise 2, basic STATA Oct 29 Week 6 Lab and review
Oct 31 Midterm Exam Nov 5 Week 7 Didactic Power, sample size
Nov 7 Didactic Survival analysis STATA lesson (Common statistical tests) Nov 12 Week 8 Didactic Correlation and Regression 10-
Nov 14 Didactic Logistic regression 11- Exercise due: Power
Nov 19 Week 9
Didactic Multivariate, Approaches to data
Nov 21 Lab Didactic : ANOVA Exercise due: Survival+ correlation and regression, logistic regression Exercise due: computer exer 3, STATA (common tests)
Nov 26 Week 10 Didactic Choices STATA lesson (data manipulation)
Holiday starts 10 p.m. today
Nov 28
Thanksgiving Holiday
Dec 3 Week 11 Lab: 10-
Dec 5 Review: Exer. due: choices Exer. due: computer 4, STATA (data manipulation) Dec 10 Week 12 Final exam
