Yes, I remember one study in particular where the participants had to look at a screen and make a decision about what they saw (for example, I agree or I disagree) and then they had to press a button with their left hand for one choice or their right hand for the other choice. Their brains were being scanned throughout the experiment and from the scans the experimenters could reliably predict almost seven seconds before the participant pressed either button, which button they would ultimately press. The participants claimed they made the decision a split second before pressing the left or right button, putting the results of the scan before when they claimed to make the decision. I found that weird, but as you say, it's very hard, if not impossible, to get the timing accurate in these studies as it relies on believing what the participant said about something that is hard to pin down.