Feb 182010

But that I could take credit for this! Unfortunately I can’t. It is gleaned from the comments section of a post on Pam’s House Blend.


  • 1 preconceived notion, strongly held
  • 7-10 random data points
  • 2-3 prior studies
  • 2-3 non-exhaustive experiments (optional)
  • & peer review

Step One – Take one preconceived notion and grasp it strongly. Bake this idea into your mind until well seared. If necessary, run around shouting it at the top of your lungs with your ears plugged and eyes shut so as to drown out any competing notions.

Step Two – With preconceived notion now permanently etched upon your thought process, begin collecting data to support it. 7-10 randomly culled data points from 2-3 prior studies should about do. In case of a lack of prior work to cherry-pick from, you may need to run several very limited experiments with no controls and little in the way of coherent methodology. Be sure to strain out any contradictory information or context, preserving only the desired bits for flavoring of your results. It is very important that you only keep that information that supports your preconceived notion.

Step Three – Take that last ingredient (peer review) and throw it out the window. Publish your “new research.” Stir scientific/professional community vigorously. Ignore any criticisms on methodology. Shout louder and people will think you’re right.

Congratulations, you’ve now baked up some junk science.

Sorry, the comment form is closed at this time.