The small sample of 25 young adult participants were asked to predict which one would randomly turn red, make a mental note of this, then wait. After one of the circles took on a crimson hue, the participants had to record via keystroke whether they had predicted correctly, incorrectly, or didnâ(TM)t have time to complete their choice.
The journal article is paywalled, so I'm relying on the linked articles for my information.
Why not have the participants enter their choice BEFORE the circle is displayed and then and automatically record whether or not their response was correct? It would keep participants honest. I think a more reasonable interpretation than "we are seeing the future because we have no free will" is that the participants simply did not have time to solidify a guess in their minds and were simply lying (even if they were doing it subconsciously) about their guess.