Lecture 11.1


I. Comparing multiple treatments in the same experiment

II. Taking the risk out of choosing the "right" levels of the independent variable so that you:

A. Don't overlook relationships among variables

B. Discover the nature of the relationship

1. Types of relationships (linear, quadratic [e.g., positively accelerating, negatively accelerating], cubic, etc.)

2. Value of determining the nature of relationship

a. Refining theory

b. Improving ability to generalize to unexplored values

3. How to choose levels

a. Number of levels

b. How far apart (proportionately)

III. Why multiple control groups are often needed

A. Confounding factors:

1. Hypothesis guessing

2. Nontreatment differences between treatment and control conditions

B. Absence of a "correct" comparison group. For example, should therapy be compared against no-treatment or against a placebo treatment?

To Lecture 11.2: An intuitive approach to ANOVA

Back to Chapter 11 Main Menu