In 2011, Leavitt and Christenfeld reported (PDF link to original paper) – to much, much, much media attention – that presenting a spoiler before a short story increased enjoyment of that short story, counter to the intuitions of many. Although this research was conducted on the usual college-students sample, the sample was very large (over 800) and the findings pretty convincing, as 11 of 12 stories got some boost in enjoyment from a spoiler. (And it turned out that Leavitt and Christenfeld got similar results reported in another paper [PDF link to original], although there they reported a few nuances to the general spoilers-improve-things finding.)

My lab reading group read and I discussed this research pretty shortly after it was published. My then-grad student, Kevin Autry, argued that the results might not generalize beyond stories that people had no prior investment in. That is, if you spoil a story that someone has no prior interest in, maybe it will make the story more enjoyable. This is not at all the same thing as spoiling a story that someone has been waiting years for (e.g., the latest Star Wars installment) or that someone has seen all but the latest episode of. This idea kicked around the lab for a while, and finally in summer 2014, my then-honors student, Michelle Betzner, and I decided to do an experiment testing whether spoiling stories after someone cared about them (i.e., in the middle of the story) was ruinous. And we also thought it was a good idea to see if Leavitt and Christenfeld’s (2011) original findings were replicable. As of that summer, there were no published reports that we could find of even an attempted replication.

The details of what we found are reported in a paper just accepted (PDF link to the accepted manuscript) for publication in the journal Discourse Processes. The short version is that we found that spoilers reduced enjoyment of stories. Our stories were different than those used by Leavitt and Christenfeld, and our spoilers were shorter and quite blunt. But it’s not probably not the shortness of the spoilers that led to the different results; Johnson and Rosenbaum – independent of my lab’s work – were doing a very similar experiment with longer spoilers, and found that spoilers reduced enjoyment of stories (press release; PDF link to manuscript). Our findings and those of Johnson and Rosenbaum are not that surprising, but in light of Leavitt and Christenfeld’s findings, they open the door to interesting questions about what kinds of spoilers may do different things to readers’ enjoyment.