Wow.

[melindadansky]

It's amazing how much trouble one can cause by flipping one little < sign.

"Oh, you wanted answers for the models that *did* work. Why didn't you say so?"

Off to stick my finger into a wall socket to re-start my heart.

Comments

[rocket_jockey]

This falls under the universal truth that math is straightforward, but arithmetic is like a bad-tempered domme waiting for an opportunity to slap you down and step on you.

Flipped that ">" during a null-hypothesis test?

[melindadansky]

That is exactly what I did--and then I dropped all the simulations that "failed".

[rocket_jockey]

I've done the same thing - not on simulation, but on data from lab experiments.