"I read this study, and it proves I'm right about depression/trauma/divorce/etc."
Here's why I primarily appeal to common sense in this blog rather than quoting research (although I do that too). Before I say this, understand that before I attended graduate school, I spent time during and after college conducting research in a biochemistry lab at Vanderbilt, the top in the world for what we were looking at. Then I worked in a psychology lab at the same institution. In addition to writing up my own work, I had to read through and completely understand a lot of papers related to that work, and speak about them intelligently with people who could make me look like a dope.
You may have found some research yourself that gives you some peace concerning your parenting and life decisions. However, one research study looking at hundreds of studies across the sciences shed some light on a serious problem with modern research: the most relevant factor in the findings of research is the theoretical framework of the researchers. CBT therapy researchers always find that CBT is the best therapy, etc. Researchers who think divorce is great find that divorce is great. In other words, people tend to find what they want to, even if they aren't consciously trying, including research into non-controversial topics. It doesn't mean the conclusions aren't true, but it means that the appropriate first reaction when reading research is to assume its inaccuracy and then be open to being proven wrong, which is the definition of the scientific method. Academia used to think this way, and had built-in safeguards against bias.
The problem with modern research is twofold, and I have seen clearly these things in-person in my lab experience. The first problem is the extreme career-survival motive to publish creative and groundbreaking research, and that depends on finding interesting new things and not having a "dud" study where you prove yourself wrong. Outright lying is extremely rare, but clearly career-related desperation influences results. The second problem is a lack of duplication of results, which guards against unconscious bias by having opposing or neutral researchers redo the study. It is no longer fashionable to try to duplicate someone's study, examining the strength of its results, although in the past this was one of the most respected activities of researchers. Do this today, and you will be a nobody with no job and no funding, and that's not what people imagine when they decide to spend 10+ years getting a Ph.D. So, modern research moves faster, but most of it is rarely read by scientists, isn't useful, and only serves to advance careers. And then journalists and others find it online and use it to support their prior conclusions. The internet makes this even worse, as you can ignore all sorts of results until you find something you like. It's like having a Formula 1 car, driving it around a go-kart track at the beach, and calling yourself world champion. And then hating anyone who says that's ridiculous and lazy, and accusing your critics of doing the thing you are doing twice as egregiously.
Serious research depends on critical readers interested primarily in learning, rather than in finding justification for their opinions--people who also read studies that disagree with the conclusions of a study. It's not always useless to quote research, but these days, if you want to be at all sure of what you're reading, you had better have a formal education in research methods, and preferably, personal experience in a research lab while you are doing that. You had better read the whole scientific paper and pay attention to the statistical analysis of the results. And that is time well-spent, especially in getting the education to do it. Even then, it's often more informative to take a critical look at your own fears and doubts regarding your opinions, and face that dragon head-on.
By the way, read up on the research related to the therapy method your therapist is using. If the method is worth anything, it will give you a lot of faith in therapy.