[♩INTRO ].

Psychotherapy — also known as talk therapy — can kind of seem like a mixed bag. You might've heard people say how awesome it can be, but just as many might say it was unhelpful, or even harmful.

For every friend or family member who swears that therapy saved their lives, it seems like there's another who says they're never going back! Most researchers do agree that psychotherapy is effective. But flawed studies make it hard to know exactly how effective it is, and it's especially tough to figure out what'll work for someone on an individual level.

As questions in psychology go, this is one of the biggest — and the trickiest. Psychotherapy comes in all kinds of shapes and sizes based on different schools of thought. And to clear up a common misconception here, most of them were not created by Freud.

Different treatments might include something like mindfulness meditation, which focuses on being aware of your experience in the moment, or cognitive behavioral therapy, which works to identify and change the thought patterns or behaviors that lead to problems. Plenty of studies have shown that many types of psychotherapy are at least somewhat effective. But every so often you'll also hear about some paper that says it doesn't help at all.

For example, a well-known article from 1983, published in the journal Behavioral and Brain. Sciences, which re-analyzed 32 studies a previous paper had shown to be effective. But this second analysis found that there was no evidence for that.

Instead, the researchers argued that the only studies that worked were the ones not done on real patients. Any study that had been done using real patients at a psychiatric hospital weren't shown to be any more effective than a placebo — a treatment that's fake but seems real. These contradictions are because, like with a lot of psychology research, the answer isn't so straightforward.

One 2017 paper really broke this down. In it, researchers from Greece, Sweden, and the UK did an umbrella review of 247 meta-analyses that measured the effectiveness of psychotherapy. A meta-analysis is like a study of studies.

It pools together all relevant quantitative data covering similar research and experiments. An umbrella review takes that one step further: It's an analysis of analyses. It's… so meta.

In total, this umbrella review covered more than 5100 trials. Of the meta-analyses the researchers looked at, 80% of them had statistically significant evidence for whatever psychotherapy treatment they were looking at — meaning, the difference in results probably didn't come from random chance. But the review showed that the majority of these analyses were littered with biases.

Only 16 of them provided what the researchers called convincing evidence, meaning they were free from bias. The others contained things like publication bias, which can happen when researchers ignore negative findings or publish only positive ones, or allegiance bias. That's when the results of research are distorted by the investigators' own preferences for certain kinds of treatments.

And it can inflate treatment effects by almost 30%. Now, this doesn't mean that we should throw all our evidence about psychotherapy out the window. For one thing, this review found bias-free results for multiple well-supported types of therapy — like how cognitive behavioral therapy, or CBT, is helpful for reducing anxiety symptoms.

By changing the way patients interpret the world, this type of therapy is effective at reducing triggers that would fuel someone's symptoms. Like, someone might interpret a quiet coworker as evidence that their office-mate — and everyone else — hates them, which could cause restlessness, nausea, or other symptoms. But CBT would help them realize that their office-mate might just be tired or shy instead.

The review also found that, when it was used along with medication, psychotherapy was more effective in putting depression into remission than medication alone. The umbrella review also found that, in addition to the 16 meta-analyses with “convincing evidence,” 20 of them had what the researchers called “highly suggestive evidence” — so even though they had some bias, their conclusions were still pretty reliable. But the results of this review could mean that psychotherapy isn't always as effective as some research has suggested.

It also means researchers need to be extra careful about how biases might affect their work, and that therapists need to look closely at the studies they're using to inform their practice. So the real challenge in this field of research isn't figuring out if psychotherapy works in general. Because we're pretty confident it can help treat all kinds of things, even if you just take into account the studies without bias.

The challenge is figuring out what will work on an individual level. Just because something works, on average, when you consider a big group of people overall, doesn't mean it'll work for you, specifically. Unlike with many other types of medical treatments, for most people there's no set formula that'll tell you exactly what type of psychotherapy you should use, or whether it'll help.

That's mostly because there are all kinds of variables involved, from the skill level of your therapist to how vulnerable you're able to be with them. It also matters how long you keep coming to therapy, because 20-50% of people quit before the time recommended by their therapist. Some of that could be because of the cost.

But people also might feel like they aren't getting better fast enough, or like their symptoms are actually getting worse. Because that can happen with psychotherapy — when you're opening up about distressing topics you've hardly ever talked about before, your state of mind might get worse before it gets better. So there are a ton of pieces to figure out when it comes to what type of therapy will work best for someone, and most of the time, therapists will look for solutions on a case-by-case basis.

But just because we don't have psychotherapy's effectiveness perfectly figured out doesn't mean it isn't worthwhile. It just means that we have a lot more learning to do. Thanks for watching this episode of SciShow Psych, and thank you, especially, to our patrons on Patreon.

We could not make these videos without you. If you'd like to support the show, head on over to patreon.com/scishow. [♩OUTRO ].