Imagine being moved from your neighborhood by the government, plopped down in a new part of town, and told that this move might change your life for the better. Or for worse. Or not at all. But there is no way of knowing in advance.

That was the premise of the Moving to Opportunity study. Researchers relocated more than 4,600 randomly-assigned families in five major U.S. cities between 1994 and 1998. The study set out to answer an important question: Would families see their economic and social prospects improve over the years if you moved them from low-income neighborhoods into higher-income neighborhoods?

This is one of the many randomized trials that Andrew Leigh highlights in his new book, “ Randomistas: How Radical Researchers Changed Our World.” Leigh says that at first, the Moving to Opportunity study suggested no strong positive outcomes for families who moved. But years later, researchers found that children under the age of 13 who moved to a lower-poverty neighborhood went on to earn $300,000 more over a lifetime.

“So it seems like moving families to low-poverty neighborhoods does have a significant effect,” Leigh says. “And this is one of those fascinating moments of ‘big data’ meets randomized trials, and we learn a whole lot more about the world.”

What is a randomized trial?

Leigh describes a randomized trial as a test where researchers assign treatments randomly. Some people get some sort of treatment and be measured against a control group — people who don’t get the treatment.

“It could be the toss of a coin, it could be randomized numbers, whether it’s a [number] generator or drawing numbers out of a hat,” Leigh says. “The key is that you want to know what the world would look like if you didn’t put in that intervention.”

One of the first known instances of a randomized trial came in the 18th century when physician James Lind was looking for a cure for scurvy. Lind separated sailors on the HMS Salisbury into groups, giving each group a different treatment. The remedies included sulfuric acid, seawater, vinegar, and oranges and lemons. Turns out, those consuming the Vitamin C-rich fruit had their scurvy cured, while those taking the other remedies languished.

But Lind couldn’t explain the reasoning behind his successful results, and instead made up “hocus-pocus,” according to Leigh.

“Respectable scientists can see that his description is essentially garbage, so they disregard his beautiful randomized trial,” Leigh says.

The Ethics Behind Randomized Trials

You’ve probably been part of at least one randomized trial at some point, whether you know it or not.

Grocery stores decide where to place products based on randomized trials. Politicians use them to figure out which campaign slogan might generate more donations. Leigh himself even used a randomized trial on Google to pick the subtitle for his book. (The worst performing option was: “How a Powerful Tool Changed Our World.”)

“Humana, Chrysler, United Airlines. A whole lot of firms are constantly tweaking and experimenting,” Leigh says. “One expert says that every pixel on the Amazon home page has had to justify its existence through a randomized trial.”

But Leigh says, though they are effective, unchecked randomized trials can spell bad news. He points to Facebook, which, a few years ago, conducted an emotional manipulation experiment. In it, they changed the content in users’ feeds to see whether positive or negative content would change people's moods. The only problem?

“[Facebook] hadn’t warned users of it, and they hadn't compensated any of them in any way,” Leigh says. “I'm all for Facebook setting up a panel of its users who've agreed to be part of the interesting cycle of psychological studies, but I don't think they should have done this experiment on users without their consent.”

Normally, though, Leigh says randomized trials are one of the most trusted methods out there. If you want proof of it, he says, you only need to see how many randomized trials pharmaceuticals go through before they’re approved.

“Most of us would not want to put our grandmother on a cancer treatment which would have shown to fail a randomized trial,” Leigh says.