Worked with a linguist years ago.
She took on the news that some non-human primate species possessed language. The primate under study had learned words, used words, had elementary syntax, and even could innovate new words according to a productive morphological pattern.
She got the video footage, the raw data. And watched many hundreds of hours of it. Collected data. This took a few years.
In the end she had the data. The original study had cherry picked. Ignore when the subject mimics a human. Ignore all the gibberish, uninterpretable utterances like "me wants banana give banana give banana me me banana banana not not banana give not me not banana not want not give give want not want" while sitting in a corner alone right after eating. All the data descended into non-significance once *all* the data were included, not just the data that fit.
Advocacy science.
The original study is still widely quoted. The disconfirmation was really quite definitive--*all* the tapes were reviewed--and that's what linguists have usually used. Although I can't vouch for the last generation.
I'll wait for confirmation of this study. Which, to be honest, depending on the amount of available (not just "used"
data and hesitancy to say something unpopular, might take years.