I said simply that we have tons of statistical data points from a multitude of peer reviewed scientific studies which suggest one thing. These studies have been done with a range of methodology. Now you can include ALL these tests together, including the ones which showed spanking had any positive element to it (and there are so few that we're talking mere single percentage points here), and the statistical data we would gather at the end would show an overwhelming consensus that spanking simply does not work.
That's why you gather all the data and evaluate its findings. You don't just look at one scientists conclusions. You don't just look at two. You look at dozens and dozens and you just keep going until you collect as much as there is out there. Then you reach conclusions based on that, no matter how uncomfortable it makes us feel. Science is not here to tell us we are right. Science is here to tell us when we need to change our minds, even if we wish the other way WERE right.
Maybe it's because you do not understand science, but nothing in your comment is at all a challenge to what I said. It doesn't matter what methodology a single test uses, because you don't use a single test. You use them all. And 'all' the tests usually end up covering the great variety of different theories and approaches people have toward studying the subject of, say, spanking. Therefore, you can analyze all this data as a group and come to firm conclusions of what would be true or not based on what the entire collective of scientific data tells us.
And then no matter what those results, you follow them, because they must be the most likely to be true. And science shows us throughout history, time and time again, that this method works. It's as simple as that. If you have come up with an alternative way of coming to objective conclusions about the world around us that works even half as well, we can perhaps continue this line of inquiry.