adrian is rad


percent vs percentage points

Filed under: — adrian @ 4:59 pm

Statistics lesson number 1, by me.

There’s a difference between percent and percentage points. People often get this wrong.

For example (bugmenot can help):

When it came to side effects, Effexor’s greatest liability was that it could cause hypertension, a side effect not shared by S.S.R.I.’s. Sussman showed us some data from the clinical trials, indicating that at lower doses, about 3 percent of patients taking Effexor had hypertension as compared with about 2 percent of patients assigned to a placebo. There was only a 1 percent difference between Effexor and placebo, he commented, and pointed out that treating high blood pressure might be a small price to pay for relief from depression.

Wrong. There was only 1 percentage point difference between the two.

3% is 50% more than 2%. (And 2% is 33% less than 3%). There is no form of comparison between the two of those percents that is only 1 percent.

(To the credit of the author, he points out that it can be seen as 50% more, though he doesn’t call Sussman’s version wrong.)

One Response to “percent vs percentage points”

  1. Colin Says:

    This is a pretty widespread problem. On ads for Amica, they say that you can get a 15% multi-car discount, which is 10% higher than the standard multi-car discount. The standard multi-car discount is 5%, so they should either say that their multi-car discount is 10 percentage points higher than the standard or that their discount is 200% higher than the standard. I would think that the 2nd option would be more attractive, but that’s not what they say…

Leave a Reply

Spam protection by WP Captcha-Free

Powered by WordPress