Your opening arguments completely disagree with the bottom quote. This isn't what you think it is. They're not saying CBT doesn't work or isn't more effective,

Your opening arguments completely disagree with the bottom quote. This isn't what you think it is. They're not saying CBT doesn't work or isn't more effective, they're saying it's not quite as effective as thought because people don't always report the failures. What's happening is that people who do a study that results in no positive benefit aren't publishing. But one study doesn't give the whole picture. There are many reasons why a singular study might show no benefit, a lot of them statistical reasons. What you need to do is look at all the studies, but the problem has been that the failing studies haven't always been published, which is bad science. So what's been happening is the published studies were: B NB B B NB B B B (6 benefit / 2 no benefit) When in fact the results were: B NB B B NB B B B NB NB (6 benefit / 4 no benefit) In science and statistics you have to report the false negatives too.