Literary fiction fans were quick to embrace news that the novels on their shelves bestowed a heightened social intelligence. The supposed link was uncovered by scientists at the New School in New York. Stories about the study were widely shared on social media platforms.

New research, however, suggests the excitement was all for naught. When social scientists at the University of Pennsylvania, Pace University, Boston College and the University of Oklahoma tried to replicate the findings of the original study, they failed to replicate the results. The conclusions of the original study simply don't stand up to scrutiny.

"Reading a short piece of literary fiction does not seem to boost theory of mind," Deena Weisberg, a senior fellow in Penn's psychology department, explained in a news release. "Literary fiction did not do any better than popular fiction, expository non-fiction and not any better than reading nothing at all."

Of course, there are a variety of perfectly valid reasons to read literary fiction. But the latest findings -- detailed in the Journal of Personality and Social Psychology -- suggests readers shouldn't do so with the expectation of an improved ability to recognize the mental states of others.

Lead researchers Weisberg and Pace's Thalia Goldstein used the reading materials from the original study -- conducted at the New School for Social Research -- and administered the same theory of mind test, the Reading the Mind in the Eyes Test, or RMET.

Weisberg and Goldstein say they don't fault psychologists for looking for a way to engage people's social intelligence, but believe scientists have a duty to always double check their conclusions.

"There's been a lot of attention to high-profile studies that show something of social importance," Weisberg said. "It would be amazing if we could put into place interventions on the basis of this study, but we really need to double check and not just rely on one lab, one study, before we go shouting from the rooftops."

The researchers did perform a separate test to see if long-term exposure to fiction was related to a high social IQ. Participants were given a list of authors, some real and some fake, and asked to verify those names which they know for certain belonged to real, live authors. Participants were penalized for selecting fake names.

Those who named the most correct authors tended to score higher on RMET, the theory of mind test. Still, the correlation doesn't prove causation.

"One brief exposure to fiction won't have an effect, but perhaps a protracted engagement with fictional stories such that you boost your skills, perhaps that could," Weisberg said. "It's also possible the causality is the other way around: It could be people who are already good at theory of mind read a lot. They like engaging in stories with people."