Getting facts right What makes good mental health research?

One of our goals with Myndplan is to support people’s effort to sift out the good stuff when they are trying to learn about psychological problems, their causes, and how to get effective treatment. At the same time we want to provide therapists with reliable information on how they can understand and help their clients. Like most people, I rely on the internet to get my facts, and the process of trying to find them has been informative. While it is very easy to get information about anything on the internet, it is not nearly as simple to find reliable guidance based on sound facts.

I’m lucky to have trained at a tech-savvy university. I also got to spend my career honing my research skills in the real world as the internet grew from a text only resource used almost exclusively by researchers to the enormous information portal that it has become today. Unfortunately, the rapid growth of the internet has meant that there is far more unreliable information than ever and makes it extremely challenging to separate the good from the bad, or in media terms, the fake from the real. Still, it is not as difficult as it may seem to cut to the chase and get at the good material if you know a few tricks. Here are a few things that I’ve learned that you might find helpful in your own searches:

1. Create an effective string of search terms

The first mistake most people make when starting to search for information is in their choice of keywords. These are the words you type into the search window when you get started, and everything else that happens will depend on which words you choose. Much of what you get back will hinge on how many and what type of words you use.

The whole is greater than the sum of its parts: Don’t just use a single word, particularly if it is the common label for a problem. While it may seem obvious to just type in the word “depression” if you are feeling blue to see what happens, the result will be very predictable – you’ll get a thousand pages of links, including a bunch of ads for things that probably won’t be too helpful. Creating a string of words that summarizes your query, such as “depression, treatment, effective” will narrow things down a bit. However, using strings of terms alone is not enough to get those pesky ads from the top of your page.

What you say is what you get: If possible, use clinical terminology rather than common terms or slang. Using depression as our example, typing in “major depressive disorder, treatment, effective” will get you very different results than when you use the simple term depression. For example, the words “scholarly articles” will appear, as well as more public and professional information sites and fewer ads.

Hint: Including the current or a recent year will limit your search to more up to date results. In psychology and psychiatry research things change quickly, so be sure you are looking at findings based on current, not outdated research.

Adding terms like scholarly, research, or study will help narrow things further to actual empirical research studies and the likes. But we’re not done yet. For most questions or problems, hundreds or even thousands of studies will have been conducted and published on the topic. To get at the good stuff, you’ll need to take some additional steps.

2. Read the Ingredients

Not all studies were created equally, but you don’t have to have a Ph.D. to spot the best research. To find the reliable papers you’ll need to understand a few key ingredients that go into a good research study, and also how you can use research about research to help your search.

Five terms you need to understand

Any good study should include a few important pieces. These include randomly sampling the population of interest, providing a placebo or other control group, and blinding the people involved in the study. You also want to know the actual size of the effect, and what the sources of bias were. If you understand what these terms mean it will help tremendously. Let’s take a quick look at each of them:

Random Sampling: Anything in mental health that’s worth reading will involve people. Who gets chosen to participate in a study matters as much as what they actually do. When researchers pick participants, it is important that this be as random as possible. Otherwise, it becomes more likely that the results can only be applied to a select group of people.

Control Group: No good study just looks at what happens when people are exposed to the experimental condition, whether this is some sort of exercise, drug, or therapy. To show real effects, it is necessary to include a group of people that don’t get the intervention – commonly called a control group. Otherwise, it is impossible to say whether the intervention caused the change or some other uncontrolled thing that wasn’t measured. Better yet, the best studies include placebo and/or sham control groups. An example of a placebo intervention is a pill that contains everything but the active ingredient. An example of a sham placebo is a pill that doesn’t contain the key active ingredient, but may have side effects from another active ingredient that are shared with the medication being studied, like dizziness or a metallic taste.

Blinding: The old saying in wartime was that loose lips sink ships. In research this saying also applies. The less the study participants know about what the expectations of the study are, the less likely they are to be influenced by these expectations. This goes for the researchers too. Good studies take pains to create double-blind conditions, where neither the people running the study nor the participants know what arm of the study they are in.

Effect Size: While you may find it surprising, just about anything can be shown to have a significant effect if you set up a study carefully. One simple rule of thumb is that the larger the group of people you look at, the more likely it is that you will find something that is statistically significant. In many cases, an effect can be highly significant from a statistical point of view, but almost meaningless from a clinical perspective. To help clarify the usefulness of findings, good studies will include something called the effect size. This is a value that estimates how strong the observed effect is, not just whether it exists.

Bias: Randomizing samples, blinding participants, and including placebo control groups are all ways of reducing bias in experimental results. These methods try to remove the effects of both the experimenters’ and the participants’ expectations so that the results reflect only the actual effects of the intervention itself. But there is another important source of bias that is frequently overlooked in science – the sort of bias that arises when researchers are in a conflict of interest. The most common example of a conflict of interest is when a researcher who is paid or supported by grants from a drug company studies the effect of one of the company’s new drugs. Bias can also occur when a scientist is testing out their own theory, rather than critically evaluating someone else’s ideas. Essentially, any time a researcher has a personal stake in the study, this introduces bias.

Hint: Typing in review or meta (short for meta-analysis) will limit your search to articles that summarize the existing research that has been done on the topic you are interested in. So if you type in major depressive disorder, treatment, effective and review or meta, you’ll get a list of articles that include mostly summaries of what all those individual studies on the topic have been showing. A fantastic source of information that scientists use for this purpose is the Cochrane Library . This is a priceless resource that more recently has built in tools to help people with little or no scientific experience to search and understand the extensive collection of information. The nice thing about Cochrane studies is that they go to great pains to evaluate the quality of research studies, including looking very closely at possible sources of bias.

3. Get the whole thing for free

If you have been trying to follow even a few of the above suggestions, you’ll probably have stumbled into a paywall. Every published research article contains an abstract – a short paragraph summarizing the purpose, method, and findings of the study. In virtually all cases, you can get abstracts without having to shell out any cash. However, when you try to look at the full text article, you’ll often encounter a paywall. To get at the content you’ll be asked to fork out an exorbitant fee. Thankfully, there are a couple of tricks that can help you get that good content for free:

Say the magic word: You can often get a free copy by typing in the name of the article you’ve found in your search bar, followed by the letters pdf. Almost every article will be available in pdf format, but pdf copies may be hidden in places like a researcher’s academic site in their biographic information, or in a post that they put up for students.

Search an open access journal: More and more journals are adopting an open access policy, which means they provide pdf downloads for free. This movement reflects a growing concern that research can only benefit society if we remove the barriers to accessing the results. Here are a few examples of how to tap into the open access world of research articles:


CORE stands for COnnecting REpositories, and is a massive search engine for finding open access articles put together in Britain. It is specifically designed to serve the general public and not just researchers. CORE is easy to use and includes some powerful filters to help narrow your search.


PLOS has quickly become one of the most popular and respected open access journals. Journals like PLOS have disrupted the traditional publication market by charging researchers for posting their work, instead of charging readers for accessing it.


The Directory Of Open Access Journals is based in Sweden and contains close to 10,000 open access journals. Like CORE it allows you to use a variety of filters in addition to your keywords.

As a final note, always remember that research articles will never be under headings that start with the google ad banner, nor will they be under a .com website, nor will they be in a newspaper or magazine.

So there you have it – with these tips you are more likely not just to find lots of relevant research on the topic that interests you, but also spot the good stuff and get a copy for free. Have a fun search!