Campbell & Stanley's threats to internal validity
Discuss Campbell & Stanley's threats to validity. To help students,
1. Make an overhead of table 8.4.
2. Realize that students will tend to have trouble understanding
regression (therefore, you may want to use demonstration 1 or
2.).
3. Realize that students will tend to confuse testing and
instrumentation.
4. Realize that students will tend to confuse maturation and
history.
5. Realize that the discussion of these threats sets up
three very important concepts:
a. correlation does not prove causality.
b. the simple experiment really is a remarkably easy way to
establish internal validity.
c. The key to experimental and quasi-experimental designs is
to show that these threats are unlikely explanations
for the obtained results.
Back to Chapter 8 Main Menu