search

Become an active member of our fallacy-discussing community (or just become a lurker!)

# Regression Fallacy

(also known as: regressive fallacy)

Description: Ascribing a cause where none exists in situations where natural fluctuations exist while failing to account for these natural fluctuations.

Logical Form:

B occurred after A (although B naturally fluctuates).

Therefore, A caused B.

Example #1:

I had a real bad headache, then saw my doctor.  Just by talking with him, my headache started to subside, and I was all better the next day.  It was well worth the \$200 visit fee.

Explanation: Headaches are a part of life, and naturally come and go on their own with varying degrees of pain. They regress to the mean on their own, the “mean” being a normal state of no pain, with or without medical or chemical intervention. Had the person seen a gynecologist instead, the headache would have still subsided, and it would have been a much more interesting visit—especially if he were a man.

Example #2:

After surgery, my wife was not doing too well -- she was in a lot of pain.  I bought these magnetic wristbands that align with the body's natural vibrations to reduce the pain, and sure enough, a few days later the pain subsided!  Thank you magic wristbands!

Explanation: It is normal to be in pain after any significant surgery.  It is also normal for the pain to subside as the body heals -- this is the body regressing to the mean.  Assuming the magic wristbands caused the pain relief and ignoring the regression back to the mean, is fallacious.

Exception: Of course, if the “cause” is explained as the natural regression to the mean, then in a way it is not fallacious.

My headache went away because that’s what headaches eventually do -- they are a temporary disruption in the normal function of a brain.

Fun Fact: Seeing a doctor can have a real effect on pain relief, even if the doctor does nothing but provide a sympathetic ear. This is known as the psychosocial context of the therapeutic intervention and is often considered part of the placebo effect.

References:

Poulton, E. C. (1994). Behavioral Decision Theory: A New Approach. Cambridge University Press.