It’s worth pointing out that the study in question was on ironic, mystery, or ‘literary’ short stories acknowledged to be high quality and subjects had to (pretend to) read the works regardless of whether they were spoiled or not.
For works of lesser quality or a different field (say, comedy), those results may not hold. If the only reason to consume a work is to get at the plot payoff, knowing what that payoff is may cause someone to avoid the work. Likewise, timing is a critical part of comedy that spoiling can often ruin.
One other flaw in that study IIRC: they only had subjects read through the stories once. In my experience, many stories can be read profitably twice: once with plot twists unknown to appreciate the mystery and suspense, and a second time (obviously with plot twists known) to appreciate the dramatic irony. The third reading then adds (relatively) little extra.
So if you have a chance to experience fiction spoiler-free, you can get twice as much out of it. But this effect wouldn’t be measured in a study where neither group goes through the work twice. You’d merely be comparing the value of “dramatic irony” (for the spoiled group) vs “mystery and suspense” (for the unspoiled), whereas a more fair comparison would be “dramatic irony” vs “mystery and suspense, then dramatic irony”.
My caveat doesn’t apply to any fiction that isn’t worth reading twice in the first place… but that may be a moot point, if like me you believe plot-dependent fiction that isn’t worth reading twice usually isn’t worth reading once either.
Right. In the shower, I also realized that this is just comparing averages- if, say, 10% of the population really hates spoilers, but the other 90% enjoys them enough to make the average for spoiling higher, it’s still sensible to put spoiler warnings as a courtesy to the 10%, because the comparison is “cost to warn vs. benefit of warning” not “spoil for everyone or spoil for no one.”
It’s worth noting that the experimental setup involved putting spoilers into the opening of an unfamiliar story—that is, the subjects didn’t know they were being spoiled. Suggests that if people enjoy stories less after learning spoilers, it may be their own fault (akin to all the studies on wine and perception).
I don’t think that’s the issue; if you look at the graphs, the standard deviation is tiny compared to the variability between stories, and in some the spoilers/no-spoilers don’t even overlap. The stats:
For all three experiments, analyses of variance revealed a significant effect of condition. (In order to control for variability between stories, we analyzed the data by comparing different versions of the same story.) Subjects significantly preferred spoiled over unspoiled stories in the case of both the ironic- twist stories (6.20 vs. 5.79), p = .013, Cohen’s d = 0.18, and the mysteries (7.29 vs. 6.60), p = .001, d = 0.34. The evocative stories were appreciated less overall, likely because of their more expressly literary aims, but subjects again significantly preferred spoiled over unspoiled versions (5.50 vs. 5.03), p = .019, d = 0.22. In all three story types, incorporating spoiler texts into stories had no effect on how much they were liked, ps > .4. Subjects also did not indicate in their free responses that they found these altered beginnings out of place or jarring.
The graphs show standard error, not standard deviation. Standard error is standard deviation divided by the square root of the sample size. It’s included on graphs to show which differences are statistically significant—it does not give a sense of the variability within a group.
Cohen’s d counts standard deviations (d=.18 means that the two means are .18 standard deviations apart), so there is actually a lot of overlap between the groups.
I agree that the small standard deviation suggests that either that doesn’t happen or the people in question are much less prevalent than 10% of the population (a number I picked because I have ten fingers). I also suspect that the mechanism roystgnr identified is stronger than the mechanism I identified.
This study isn’t set up to differentiate between people, which is what we would need to make a warning policy.
(I had an erroneous statement about the sample size here, which I’ve deleted.)
Hmm. That looks like a memory error on my part, as rereading it I don’t see what I thought the n was (I remembered ~40). I think I saw 30 subjects, failed to multiply by 24, and it got fuzzed with the passing of time.
This is, I think, a general problem with many of the studies we rely on. We learn a lot about the averages, but often I find that useless—I want to know more in-depth information about the subgroups, outliers, etc. As seen here and in the linked article, these same studies are often misinterpreted for this reason. Perhaps it’s worth a post? Especially if someone is more familiar with the pertinent methodology than I.
Upvoted for the spoilers don’t spoil stories link—that is incredibly useful. I can now write reviews without worrying about hiding plot details.
Please don’t do that—that is not the correct conclusion to reach from the study.
Please have the courtesy of warning in advance of spoilers when you’ll be writing a spoiler review.
It is extremely obnoxious to not allow the potential readers their own choice of whether they want to be spoiled or not.
It’s worth pointing out that the study in question was on ironic, mystery, or ‘literary’ short stories acknowledged to be high quality and subjects had to (pretend to) read the works regardless of whether they were spoiled or not.
For works of lesser quality or a different field (say, comedy), those results may not hold. If the only reason to consume a work is to get at the plot payoff, knowing what that payoff is may cause someone to avoid the work. Likewise, timing is a critical part of comedy that spoiling can often ruin.
One other flaw in that study IIRC: they only had subjects read through the stories once. In my experience, many stories can be read profitably twice: once with plot twists unknown to appreciate the mystery and suspense, and a second time (obviously with plot twists known) to appreciate the dramatic irony. The third reading then adds (relatively) little extra.
So if you have a chance to experience fiction spoiler-free, you can get twice as much out of it. But this effect wouldn’t be measured in a study where neither group goes through the work twice. You’d merely be comparing the value of “dramatic irony” (for the spoiled group) vs “mystery and suspense” (for the unspoiled), whereas a more fair comparison would be “dramatic irony” vs “mystery and suspense, then dramatic irony”.
My caveat doesn’t apply to any fiction that isn’t worth reading twice in the first place… but that may be a moot point, if like me you believe plot-dependent fiction that isn’t worth reading twice usually isn’t worth reading once either.
Right. In the shower, I also realized that this is just comparing averages- if, say, 10% of the population really hates spoilers, but the other 90% enjoys them enough to make the average for spoiling higher, it’s still sensible to put spoiler warnings as a courtesy to the 10%, because the comparison is “cost to warn vs. benefit of warning” not “spoil for everyone or spoil for no one.”
It’s worth noting that the experimental setup involved putting spoilers into the opening of an unfamiliar story—that is, the subjects didn’t know they were being spoiled. Suggests that if people enjoy stories less after learning spoilers, it may be their own fault (akin to all the studies on wine and perception).
I don’t think that’s the issue; if you look at the graphs, the standard deviation is tiny compared to the variability between stories, and in some the spoilers/no-spoilers don’t even overlap. The stats:
The graphs show standard error, not standard deviation. Standard error is standard deviation divided by the square root of the sample size. It’s included on graphs to show which differences are statistically significant—it does not give a sense of the variability within a group.
Cohen’s d counts standard deviations (d=.18 means that the two means are .18 standard deviations apart), so there is actually a lot of overlap between the groups.
I agree that the small standard deviation suggests that either that doesn’t happen or the people in question are much less prevalent than 10% of the population (a number I picked because I have ten fingers). I also suspect that the mechanism roystgnr identified is stronger than the mechanism I identified.
This study isn’t set up to differentiate between people, which is what we would need to make a warning policy.
(I had an erroneous statement about the sample size here, which I’ve deleted.)
Small n? They used 819 subjects—that’s bigger than pretty much any psychology cited on LW!
Hmm. That looks like a memory error on my part, as rereading it I don’t see what I thought the n was (I remembered ~40). I think I saw 30 subjects, failed to multiply by 24, and it got fuzzed with the passing of time.
Thanks for the correction!
This is, I think, a general problem with many of the studies we rely on. We learn a lot about the averages, but often I find that useless—I want to know more in-depth information about the subgroups, outliers, etc. As seen here and in the linked article, these same studies are often misinterpreted for this reason. Perhaps it’s worth a post? Especially if someone is more familiar with the pertinent methodology than I.
I agree. He should have started a post on that. The SMBC comic isn’t nearly as interesting.
I wonder if the stories were universally enjoyed more, or just better on average. Perhaps some people do prefer not being spoiled.
I could have sworn I’d seen a post on LW about that story. Maybe I saw it linked from planetrationalist or something instead.
I did mention it recently with a PDF link.
I linked it on Google Plus a while ago.
It’s possible I missed it.