I don't quite follow this: From the table, in the 5 years 2010 to 2014, it shows 32 trials, 7 of which were behavioural, meaning 25 non-behavioural. From the table, in the 4.25 years 2015 to date, it shows 23 trials, 2 of which were behavioural, meaning 21 non-behavioural. So given there is still 9 months of 2019 to go, it suggests to me that non-behavioural will have likely dropped very little, if at all. And that behavioural research has indeed dropped off, the far more likely reason being that the penny is genuinely dropping that it is a scientific dead end, as PACE effectively proved.
The idea of using this metric seemed really bizarre to me. Who cares how many poorly designed trials of ineffective treatments there are? Thanks to John for providing the figures. It makes their value seem even more questionable.
It's a fair point and perhaps I could have said a little more on that. I have now added something to the post:
Yes, in my post I just wrote down as .25, whereas in my head I was meaning a quarter, as a rough approximation in terms of months. But of course given I stated it to 2dp I sort of blew it as an approximation .
Defending cherry-picked research by cherry-picking some more. On brand. They must be running out of place for all the cherry pies they'll have to cook if they don't want to spoil all those finely-picked cherries.
Time to call in Special Agent Dale Cooper for some darn fine cherry pie https://giphy.com/gifs/trippy-twin-peaks-smoke-PVLb0qOwbe0y4 Edited to try and get the gif to appear ....