Many memory researchers believe in the so-called “end effect.” When we evaluate something that happened to us, the thinking goes, the final moments of it are disproportionately influential. Maybe the movie, as a whole, was kind of a bore, but that twist ending was brilliant. Or maybe the vacation was underwhelming because of the cramped hotel room, but the final day featured an astounding volcano hike. In these cases, when we look back on the movie or the trip as a whole, we’ll do so more favorably than if we simply averaged out each “moment” of the experience. If the end effect is true — and it’s a well-respected idea in the literature — it could have important ramifications for how we structure vacations and other experiences.
But a new study suggests it might not be. Writing in the Journal of Experimental Psychology: General (unlocked draft version available here), doctors Stephanie Tully of the University of Southern California and Tom Meyvis of New York University present a set of experimental findings that, taken together, suggest the end effect may not be real.
The pair ran a series of carefully controlled experiments designed to isolate the end effect. In one, for example, they had a bunch of undergrads listen to an unpleasant noise for a period of time (God bless undergrads, those hardy experimental subjects). In the control condition, the students simply listened to the annoying noise. In another, a period of less annoying noise was added to the end, and in a third it was added to the beginning.
The end effect would predict that those undergrads in the group with the less annoying end would rate the noise as more tolerable, but that wasn’t what happened — instead, there was no statistically significant difference between the ratings of the two non-control groups, and both, as predicted, evaluated their noises as less annoying than did the members of the control group. The researchers consistently found these sorts of results over the course of their seven studies, including one that took things out of the lab — the researchers measured participants’ enjoyment of a 12-obstacle fun run. In the end, Tully and Meyvis concluded that while people might overweigh the end of an experience in a limited set of specific instances, their experiments offered evidence to support the idea that people inherently do this.
This all goes against a fair amount of past research. Why have so many other researchers published findings supporting the end effect? One possibility is, well, math. Let’s say you watch a three-minute video of, say, a sea lion cuddling with a dog and are asked to rate how much you enjoyed it on a scale of 1 to 5. Imagine you gave the first and second minutes a 3, and the third minute — the stirring climax where the sea lion and canine become best friends — a 5. If you averaged out the minute-by-minute scores, you’d give the film a 3.66. But let’s say, as many researchers believe, it’s the last minute that’s the most important, meaning you overweigh it (count it more than the other two minutes). In that case, what would you give the video? A 4? A 4.3? That’s not that far off from a 3.66, and most studies are more complicated than this rather oversimplified example. So, because it’s not easy for humans to quantify enjoyment, some experimental results from the past may have been misinterpreted by researchers — the numbers seemed to support the existence of the end effect, but what was really going on was that a “good” or “bad” ending simply got factored into the overall average.
That said, one study can’t fully overturn the idea of an end effect, and the researchers do explain that there are plenty of important instances in which the end of an experience does matter more:
Endings are likely… over-weighted when the last part of an experience is particularly meaningful. For instance, evaluations of goal-directed experiences may be particularly affected by the end of the experience (Carmon & Kahneman, 1996) because the end often determines whether the goal has been met. The end may also be particularly influential for narrative experiences, such as watching TV shows (Hui et al., 2014), because the end of an episodeoften provides a meaningful resolution.
Finally, endings can also have a disproportionate impact through recency effects. Specifically, for experiences that are long and varied (e.g., a year-long trip around the world), people may simply be unable to remember many parts of the experience because of memory constraints. In that case, basic memory research suggests that the overall evaluation may be disproportionately influenced by both the beginning and end of the experience, rather than what happened in the middle[.]
Memory is complicated, and — to invoke the old refrain — more research is needed.