Sometimes people will distinguish between "replicating a result" and "reproducing an experiment" – with the differences more prominent when you are dealing with

Sometimes people will distinguish between "replicating a result" and "reproducing an experiment" – with the differences more prominent when you are dealing with straight data. Replicating a result simply means you confirm that the results aren't made up – you start with what they say they started with, you get the result they say. Reproducing an experiment might mean working from their conclusions, not trying to repeat every detail, and seeing if you come up with the same result – you might not if you mess up the experiment, but you might also confirm the interpretation of the result because you've eliminated aspects that the theory considers insignificant. With results derived strictly from data this is a clearer line, as repeating the process exactly is quite possible (the spreadsheet in this case is the "process"). This might seem trivial, but it's pretty common that even repeating the purported process is very hard. From a techy point of view this post helped me see the distinction: http://ivory.idyll.org/blog/replication-i.html http://ivory.idyll.org/blog/replication-i.html