|
|
Nemours Biomedical Research Statistics Course
October 2008
Overview
The validity and reliability of any research depends on the adequacy of the study design, accuracy of the data collected, appropriateness of data analysis and interpretation of the results. Statistics deals with all these issues. To obtain a valid and reliable result from a research project, a sound statistical principle and method must be applied over the entire life cycle of the project. Thus, from project inception with a protocol design, through sample size selection, endpoint selection, participant recruitment, randomization and treatment allocation, measurement, analysis, and interpretation, statistical ‘best practice’ must be observed.
This course is designed to provide an overview of statistics to participants with focus a on study design, variable measures, data analysis and interpretetion. The topics to be covered include descriptive statistics, hypothesis testing, correlation, regression analysis, analyse of variance, repeated measures, logistic regression, survival analysis, parametric and nonparametric tests. In class, SPSS/MS-Excel will be used for the purpose of data analysis and output interpretation. Participants will be expected to gain knowledge in designing clinical studies, data analysis, presentation, and interpretion of results.
Goals
The primary goal of this course is to teach the basics of research methodology, statistical analysis, data interpretation and presentation of results. A second goal will be to enhance a collaborative and multidisciplinary team approach to problem solving through statistical reasoning.
Course Director
- Stephen Lawless, M.D., M.B.A. (Nemours Quality & Safety)
Course Coordinator
- Tim Bunnell, Ph.D. (Nemours Biomedical Research/Bioinformatics Core)
Instructors
- Jobayer Hossain, Ph.D. - Biostatistician (Bioinformatics Core)
- Laurens Holmes Jr., Dr.P.H., Ph.D. - Epidemiologist (Orthopedics)
Eligibility and Class Composition
This course is available to Nemours Associates, primarily in research and clinical positions. The class size will be limited to 16 participants. Approval of department head will be needed for participation. There will be a pre-course screening. Class work will involve both individual and group assignments.
Prerequisites
Knowledge of basic statistics and statistical software.
Assignments
- 3 Take-home assignments
- 1 Take-home final exam/assignment
- Assigned in class 8
- Final
- Return for final comments
Data sets
Handouts
Schedule - Click on the class 'Date' for PowerPoint notes
Week
|
Date
|
Topics
|
Instructors
|
| 1 |
10/02/2008
|
Course Overview: Research Design and Methodology - Study Design,
Objective, Variables, Analysis techniques, etc
|
Dr. Hossain
|
| 2 |
10/09/2008 |
Study Designs Epidemiologic (Observational) and
Clinical Trials (Experimental)
|
Dr. Holmes
|
| 3 |
10/16/2008 |
Descriptive statistics and data visualization:
Measures of central tendencies, dispersion and
shape of the distribution
|
Dr. Hossain
|
| 4 |
10/23/2008 |
Inferential Statistics: Inferences, test of hypothesis, type I
and type II error, level of significance, power, confidence
interval, p-value, etc
|
Dr. Hossain
Dr. Holmes
|
| 5 |
10/30/2008 |
Parametric and non-parametric tests: Normal test, t-test,
Chi-square test, F-test, sign test, signed rank test,
rank test, Mann-Whitney U-test
|
Dr. Holmes
Dr. Hossain
|
| 6 |
11/06/2008 |
Correlation and Regression analyses
|
Dr. Hossain
|
| 7 |
11/13/2008 |
Logistic Regression (odds ratio), and Survival
analysis (hazards ratio)
|
Dr. Holmes
|
| 8 |
11/20/2008 |
Analysis of Variance (ANOVA), Repeated Measures ANOVA
|
Dr. Hossain
|
|