Yes, I'm talking about the CGI figures, where patients reported a significant improvement. (There are no separate figures for a 'little better'... So I'm assuming that the authors considered these results to be an insignificant change, which is why they haven't included the figures separately, but they are combined with 'no change' and a 'little worse'.) I'm not really commenting on what the researchers were bound to report... But they appear to me to have spun the figures because they haven't reported them as a comparison to the control group... So i think it's important for us to pick up on this. I haven't looked at the other figures yet... But from what you have said, it looks like only 15% of patients (6/10 - 4.5/10 = 1.5/10 = 15%) were helped by CBT or GET over and above the control group, using the other methods of measurement?