spurious correlations

I assume you all know 90% of statistics are made up, but I was delighted to find some great examples of co-related variables. That aren’t.

I can’t remember now, but when I studied stats I seem to recall an R value of over 0.9 or so being suggestive of a high correlation (but not causation). I guess it depends too on sample size etc. etc.

Anyway here is great website with enough doubtful correlations to calm down even the most ardent pointy haired boss on a ‘data driven mission’ to do something retarded.

Have fun, and don’t eat too much cheese.

One Response to “spurious correlations”

  1. Stephen Allen Says:

    Your examples are good fun. Unfortunately some claimed correlations in financial audit can have devastating and wholly unwarranted effects.

    The specious interpretation of what is “the population” and what are findings of fact associated with that population was used in the now defunct Audit Commission to limit central Government grants to local authorities in relation to the old Housing and Council Tax Benefits. The perceived professionalism of the auditors within the Commission made it very difficult to challenge their interpretation, even when they were quite clearly wrong.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: