> I just throw out papers that don’t publish datasets and code
There was an interesting research paper that claimed you could up your IQ with a computer game called dual n back and a lot of paper tried to replicate it, but my mind was blown when I realized none of them used the same code. None of them actually shared the code they used for their research and then they all claimed they were testing the same thing when they obviously weren't.
To me, refusing to share the source code to a program that could be used to replicate your research seems like a big middle finger to science itself. It shows a total disregard to study replication.
There was an interesting research paper that claimed you could up your IQ with a computer game called dual n back and a lot of paper tried to replicate it, but my mind was blown when I realized none of them used the same code. None of them actually shared the code they used for their research and then they all claimed they were testing the same thing when they obviously weren't.
To me, refusing to share the source code to a program that could be used to replicate your research seems like a big middle finger to science itself. It shows a total disregard to study replication.