The Champ has been used in experiments to see if depressed people are more likely to cry than non-depressed people (they aren’t). It has helped determine whether people are more likely to spend money when they are sad (they are) and whether older people are more sensitive to grief than younger people (older people did report more sadness when they watched the scene). Dutch scientists used the scene when they studied the effect of sadness on people with binge eating disorders (sadness didn’t increase eating).

The story of how a mediocre movie became a good tool for scientists dates back to 1988, when Robert Levenson, a psychology professor at the University of California, Berkeley, and his graduate student, James Gross, started soliciting movie recommendations from colleagues, film critics, video store employees and movie buffs. They were trying to identify short film clips that could reliably elicit a strong emotional response in laboratory settings.

It was a harder job than the researchers expected. Instead of months, the project ended up taking years. “Everybody thinks it’s easy,” Levenson says.

Levenson and Gross, now a professor at Stanford, ended up evaluating more than 250 films and film clips. They edited the best ones into segments a few minutes long and selected 78 contenders. They screened selections of clips before groups of undergraduates, eventually surveying nearly 500 viewers on their emotional responses to what they saw on-screen.

Some film scenes were rejected because they elicited a mixture of emotions, maybe anger and sadness from a scene depicting an act of injustice, or disgust and amusement from a bathroom comedy gag. The psychologists wanted to be able to produce one predominant, intense emotion at a time. They knew that if they could do it, creating a list of films proven to generate discrete emotions in a laboratory setting would be enormously useful.

Scientists testing emotions in research subjects have resorted to a variety of techniques, including playing emotional music, exposing volunteers to hydrogen sulfide (“fart spray”) to generate disgust or asking subjects to read a series of depressing statements like “I have too many bad things in my life” or “I want to go to sleep and never wake up.” They’ve rewarded test subjects with money or cookies to study happiness or made them perform tedious and frustrating tasks to study anger.

“In the old days, we used to be able to induce fear by giving people electric shocks,” Levenson says.

Ethical concerns now put more constraints on how scientists can elicit negative emotions. Sadness is especially difficult. How do you induce a feeling of loss or failure in the laboratory without resorting to deception or making a test subject feel miserable?

“You can’t tell them something horrible has happened to their family, or tell them they have some terrible disease,” says William Frey II, a University of Minnesota neuroscientist who has studied the composition of tears.

But as Gross says, “films have this really unusual status.” People willingly pay money to see tearjerkers—and walk out of the theater with no apparent ill effect. As a result, “there’s an ethical exemption” to making someone emotional with a film, Gross says.