Comment by Tloewald

It’s worse than that:

4) researcher is a serial p-hacker who even boasted about the practice

5) reason given for the small sample (most recipes had changed) was a lie and sample appears to have been cherry-picked