This is how I normally look at things. If I can't easily replicate an experiment, then it's very likely to be wrong.

Sadly, it's pretty rare (and exciting) when you can easily do something based on the methods in someone's paper