Laura Germine, PhD
Laura Germine, PhD, uses web-based tests to learn more about cognitive processes

We’ve all seen those quizzes on the web and in our social media feeds: Which 90s TV star are you? What character would you be in the Harry Potter universe?

Normally, we see them as a fun and inconsequential distraction, a way to pass the time while we’re waiting in line at the store or unwinding at the end of the day.

It turns out, however, that web-based quizzes can be a powerful research tool.

While they may not be able to tell you what celebrity you resemble or what your spirit animal would be, a team of researchers led by Mass General’s Laura Germine, PhD, has been using web-based testing for more than a decade to gain valuable scientific insights into the mysteries of human cognition.

Germine, an investigator in the Mass General Department of Psychiatry, is a pioneer in the growing field of citizen science, a process where large numbers of people contribute to science—usually online—by volunteering to participate in experiments posted by researchers.

By posting cognitive assessment tests on her website,, Germine has been able to gain access to a much larger pool of research subjects than would be possible with a traditional in-lab research study.

She has also been able to capture the types of subjects who have been difficult to recruit into research, such as middle-aged subjects who are typically at work or with children when most research studies are conducted in the lab.

“It’s nice to be able to have an experiment where instead of having 40 to 50 to 100 people, we get 1,000 to 5,000 to 10,000 people,” Germine explained in a recent interview. “It makes for science that is more statistically powerful. In a traditional trial it’s usually not feasible to get that many people in the door. It’s just never going to happen. If you could, it would be at an enormous expense.”

Germine’s work has led to some fascinating insights into the way the mind works. A recent study demonstrated that different mental functions in the brain achieve peak performance at different ages, challenging the long-established notion that the brain’s functioning peaks as a whole when we are young, and then goes into a slow and steady decline as we age.

Working with sets of identical and non-identical twins, Germine has also shown that a person’s sense of attractiveness in others is based more on life experiences than on genetics.

The Beginning—An Online Test Goes Viral

In a way, Germine feels like the direction of her research program has been driven largely by serendipity. Her research platform arose out of an effort to create tools that patients and participants could use to learn more about themselves, and not necessarily for research.

As a student in London back in the early 2000s, Germine was studying a disorder called face blindness, or prosopagnosia, a condition where it is difficult to recognize familiar faces—even family members—when seen out of context.

At the time, it was assumed that the vast majority of people who had this condition had suffered some kind of brain damage, such as that which occurs due to a stroke.

However, there was a small but active community of people who claimed they had this disorder without having ever suffered brain damage, even though their doctors told them it wasn’t possible. This community began sharing their experiences with each other and their numbers grew.

The standard neuropsychological tests for people with brain damage are easy to cheat on if you have no brain impairments, Germine explained, so the lab team created a more nuanced test that could identify the condition in a more reliable way.

When news of this new test spread, many of those who claimed to experience face blindness without brain damage asked to come into the lab to be tested. There was more interest, in fact, than the lab could handle.

So Germine created a web-based version of the test and put it online. At the end of the test, users would receive their results based on the same criteria that researchers had used to diagnose impairment in the lab. The team then sent the a link to the site out to a few people who had made inquiries.

A couple weeks later, the site crashed. To the surprise of Germine and her colleagues, it turned out that there were thousands of people taking the test, far more then the few they had initially emailed.

It turns out that in addition to those who claimed to have face blindness, everyday people were also taking the test because they had been sent the link by someone and wanted to see how good they were at facial recognition.

“Everyday people thought it was fun and liked that it was science-based.” Germine says. “They wanted to see how good they were. They wanted to help and learn a little about themselves.”

Given the overwhelming success of this test, Germine started to think it might be possible to collect viable data online. It might seem obvious in retrospect, she acknowledges, but at the time it was an evolving notion.

Building a Website

When Germine started graduate school in the Harvard Psychology Department in 2008, she created as a way to collect data in large online studies. Once again, she launched the site in a low key form and was surprised when it experienced a similar spike in traffic once going live. Over the past eight years, it has collected information on more than 1.5 million participants. 

The idea behind the site is to create experiments that participants will actually enjoy taking part in, Germine said.” You are not paying them, but you are giving them a reward in terms of sharing some of the data you are collecting for yourself.”

Germine says the vast majority of people are taking the tests because they want to know about themselves, they are interested in the mind and the brain, or just looking for something fun. Some are taking the test because they are worried about cognitive deficits, and are using the site as a substitute for or enhancement of formal testing. Others have just started taking medication and are wondering if it is affecting their cognition.

Validating the Data

It hasn’t always been easy for Germine to convince her scientific peers about the quality of data derived from online sources. When she first started her work in the mid-2000s, many academics argued that the people who take tests on the internet might be purposefully, maliciously trying to interfere with your work.

But for most people, the concept of the internet has evolved, she says. “Now we understand that the majority of people online do not have malicious intent. Now that the internet is a place where we live, it’s not so controversial.”

A lot of the early work that Germine and her team did was to establish that the data they received online was comparable to the results they were getting in the lab. “In terms of basic quality metrics, how reliable was this data?”

What they learned was, if anything, the web data ends up being a little higher in quality. This may be because participants are electing to take the test because they have an interest in receiving accurate results.

“In a traditional setting where you are paying someone for participating, some people are doing it for the money. They don’t care about the experiment. They sit there and hit keys until they get their money and go home.”

What the Future Holds

In the eight years since she began her website, Germine and her team are still finding new ways to use online platforms to answer scientific questions. “I feel like it’s a big space to explore with a lot of potential and we’re just starting to do that, and it’s exciting.”

There is a large unmet need, especially when it comes to integrating cognitive research assessments with clinical tools and patient care, Germine says. With all of the different apps and gadgets that can be used to gather health information about patients remotely, there could be a lot of opportunities to combine online testing with physiological data gathered in real time.

“I hope that it will only mature and it will be something that is good for healthcare, good for patient sovereignty and good for science.”

Back to Top