Regression Fallacy

1
Regression Fallacy
2
Pre-test, post-test

• A preschool program attempts to boost childrens
  IQs.
• The children are tested when they enter the
  program (pre-test) and again when they leave
  (post-test)
• On both occasions, the scores average out to 100
  with a SD of about 15.
• The program seems to have no effect.

3
Program effects

• On a closer look, the children who were below
  average on the pre-test had an average gain of
  about 5 IQ points.
• Conversely, those children who were above average
  on the pre-test had an average loss of about 5 IO
  points.
• Does the pre-school program equalize intelligence?

4
The regression effect

• We cannot expect the children to score the same
  on both tests. There will be differences. These
  differences will make the scatter diagram for the
  test scores into a football-shaped cloud. The
  spread along the SD line makes the bottom group
  come up, and the top group come down. This is the
  regression effect.

5
The regression fallacy

• The regression effect is the fact that in
  virtually all test, re-test situations, the
  bottom group on the first test will on average
  show some improvement on the second test, and the
  top group will on average fall back.
• The regression fallacy is thinking that this
  regression effect must be due to something
  important, rather than just spread around the SD
  line.