mdsc 643.02 biostatistics ii
DESCRIPTION
MDSC 643.02 Biostatistics II. Gordon Hilton Fick [email protected] www.ucalgary.ca/~ghfick 220-6939. Teaching Assistants. Jian Kang: Office hours and some large group sessions Shelly Vik: Marking and grading. Stata. First: Phone Stata Corporation: 800-248-8272 - PowerPoint PPT PresentationTRANSCRIPT
Teaching Assistants
• Jian Kang:
• Office hours and some large group sessions
• Shelly Vik:
• Marking and grading
Stata
• First:• Phone Stata Corporation: 800-248-8272• Intercooled Stata 8 is recommended• Perpetual license is the best deal: US$134• You want ‘Method 3 Grad Plan’• Then:• Vicki Stagg will contact you• [email protected]• 220-7265
Appointments to see GHF
• Crystal Elliott
• 220-4288
• Monday afternoons are best
• Other days and/or times are possible too
Assignments
• Interpretation most important
• All Stata analyses must be explained
• Define all terms in context
• Use Equations editor to make symbols answermybx unadj *
1̂
Prerequisite Course
• MDSC 643.01: Biostatistics I
• B+ or better
• Within the past 3 years
• See GHF if you have other preparation
Prerequisite material
Interpret accurately and completely:• boxplots, scatterplots, line plots, axes, units,
titles• mean versus median: when?• data transformations• tests and confidence intervals• tables: FET versus tests• means: t-tests, analysis of variance• matched analysis
2
Computing Background
• Windows/Mac/Linux architecture
• Website: links, left versus right clicks
• Stata : Rabe-Hesketh & Everitt
• : Chapters 1 & 2
Projects
• Grouped In pairs
• Power Point presentations
• To display biostatistical content
• 10 minutes for each group
Course Objectives
• Health Studies: From idea, to design, to data collection, data analysis and interpretation
• Linking classical analyses to model-based analyses• Using software: spreadsheet/database to statistical
analysis• Linear regression and Logistic regression: choosing,
assessing, interpreting• Introducing other regression models: conditional logistic,
proportional hazards and others• Using models to assess potential confounders and/or
modifiers