Multiple Regression Analysis
Study considering systolic blood pressure (SBP), body size
(QUET), age (AGE) in a hypothetical sample of 32 white
males over 40 years old from a town called Angina.
Suppose we believe that SBP
Contingency Tables and
Measures of Association
A contingency table is a table used to display data that can be
classified by two (or more) categorical variables. It displays
the frequency distribution in tabular format.
Introduction to Regression Analysis
Regression analysis is a statistical method used to
assess relationships between an outcome or
dependent variable of interest (Y) and one or more
predictors or independent variables (X1
Correlation is made up of co (together) and relation.
Corrleation between random variables is positive when values either
increase together or decrease together.
Correlation is negative when one value tends to increa
Lecture 1: Review of
Some Terms and Concepts Related to
Methods of statistical analysis used to assess the
relationship between a set of variables (ie. Between an
outcome variable (response) and predictor variab
Dummy (indicator) Variables
Dummy (indicator) variables take on a finite number of values.
They index categories of a nominal variable. Values of the variable
are not meaningful, but represent categories of interest.
Intro to Logistic Regression
Logistic Regression is used when the dependent variable (outcome under
consideration) is dichotomous (binary).
Interest is usually in modeling the probability of some event, given the value o
Confounding and Interaction
Goals of regression analysis:
1) Finding a model to fit observed data well and
accurately describe future observations
2) Accurately describe the relationship between one or
more regression coefficients and outcome
Techniques to check assumptions and assess
accuracy of computations for multiple regression
Simple Approaches to Diagnosing Problems in Data
Familiarity with basic characteristics
Selecting Best Regression Equation,
Selecting Best Regression Equation
General Problem: One response variable (Y) and a set of K predictor
Goal 1: Determine which of the K predictor variables best predict Y.