← Back to context

Comment by jrockway

14 years ago

Yeah, I remember a number of labs where I didn't get the data that was expected and had to do the experiment again. Knowing the expected result, producing excellent data was easy; instead of a painstaking data-collection process, I outsourced that to a "function" with "noise". Hey look, r=0.999!

I understand this is the same process by which the top quark was discovered.

Hey look, r=0.999!

I had to grade physics labs one semester -- this was always an instant tell that students were making shit up. :)

  • I wish I'd even once had students who understood the experiment well enough to fabricate data.

  • always an instant tell that students were making shit up

    Including the correlation coefficient!