Empirical studies of trigger/content warnings

General Off-Topic Discussion
1

I don’t use it flippantly simply because I know to do so can easily annoy people, and it’s not a case where that’s much of an impingement on my behaviour.

I feel this is quite decent of me as I advocate against blanket use or blanket belief in trigger warnings, and I say this, and why, whenever people are advocating them. I regularly revisit the state of research on them. There are theories both ways, but empirical evidence is so far one way.

The 2020 introduction to a typology study (trying to sort how warnings are presented, much like our own recent topic on intfiction) says:

More recent studies have experimentally investigated the impact of content warnings, especially in educational settings. A randomised study found participants with no trauma history (n = 133) who received warnings before reading passages with disturbing content reported more anxiety than those not receiving (n = 137) warnings, suggesting warnings can undermine emotional resilience [15]. The same authors replicated this finding with a college student sample (n = 462) [16], and also showed in a randomised study of trauma survivors (n = 451) that content warnings inadvertently reinforce the centrality of trauma experiences to identity [17]. A 2019 meta-analysis confirmed this finding that content warnings are associated with increased anxiety and negative mood [1]. Meta-analyses of a series of studies involving students and internet volunteers, with and without a trauma history, found mainly neutral or slightly negative impact of content warnings, leading the authors to conclude that such warnings are neither meaningfully helpful or harmful [18].

The summary of the 2023 meta-analysis of other studies:

Overall, we found that warnings had no effect on affective responses to negative material or on educational outcomes. However, warnings reliably increased anticipatory affect. Findings on avoidance were mixed, suggesting either that warnings have no effect on engagement with material or that they increased engagement with negative material under specific circumstances. Limitations and implications for policy and therapeutic practice are discussed.

The highlights of the 2018 empirical study:

Trigger warnings increase peoples’ perceived emotional vulnerability to trauma.
Trigger warnings increase peoples’ belief that trauma survivors are vulnerable.
Trigger warnings increase anxiety to written material perceived as harmful.

The 2020 Helping or Harming study acknowledges in its conclusions:

Public arguments regarding trigger warnings have been politically charged, complex, and data-poor. Recent research on trigger warnings can importantly inform or perhaps even settle some of these debates. The research suggests that trigger warnings are unhelpful for trauma survivors, college students, trauma-naïve individuals, and mixed groups of participants (Bellet et al., 2018, 2020; Bridgland et al., 2019; Sanson et al., 2019). Given this consistent conclusion, we find no evidence-based reason for educators, administrators, or clinicians to use trigger warnings.

Whether trigger warnings are explicitly harmful is less clear. We found evidence that trigger warnings increase the narrative centrality of trauma among survivors, which is countertherapeutic (Boals & Murrell, 2016).
…
There are contextual limitations on the research that you will find discussed in each piece. Nevertheless, all the best science we have so far keeps indicating that trigger warnings are not helping, and may be harming. Not just to people who think they’re helping themselves and people they know, or who heed such warnings, but in fact to anyone who reads the warnings. At best, with a negligible effect. This is not to do with any cultural alignments, just what we have tested. I keep rechecking the research base as I’m particularly interested in mental health therapies and theories.

-Wade

9 Likes
2

The lived experiences of people I know and care about who find trigger warnings useful will always carry more weight to me than research findings.

7 Likes
3

Regarding the study about content warnings: Off course the anxiety is statistically higher after a CW announcing something potentially harmful. I am privileged that I have no need for a trigger warning. But I know personally at least one person who has experienced something extremely traumatical, and so I know CWs are good. Really.

6 Likes
4

My hot takes based on these studies:

I can see how content warnings could increase feelings of suspense/dread. If you are working in a horror genre, these could actually be a useful tool to create mood.

Seems like the participants didn’t have the option to not read the passage? Avoiding material you don’t want to read seems like a really good use case for content warnings, but this was not tested. So, I think the jury’s still out on content warnings.

5 Likes
5

This hits the nail on the head, imo. I have very particular things that set me off, and if I’m given the choice, I will simply not play games/watch shows/read books that feature those particular things.

Those studies are specifically focused on educational settings, where students are expected to read the material that’s presented to them by the instructor, and don’t really have a choice in the matter. That’s not something that’s especially common in IF—if a particular IFComp entry warns me that its content could set me off badly, I can simply not play that one. There are dozens and dozens of others I could play instead.

The same is true for content warnings in general, in cases that have nothing to do with trauma triggers. Mathbrush, off the top of my head, has said several times he doesn’t enjoy IF with explicit sexual content, so he just chooses not to play it. And there’s no real drawback to him for that (it’s not like he’s going to fail a class because he didn’t play First Contact), nor is he in a setting where he’s expected to build emotional resilience (most of us are on this forum for fun, not for education or therapy).

Tl;dr those studies are focused on a particular context that I don’t believe is relevant to IF in general.

17 Likes
6

Having recently graduated, I thought it was worth mentioning that in my experience, professors are very often willing to work with students when it comes to potentially triggering situations. Not always, and the extent to which they were accommodating varied, but generally, professors were people who had compassion for a student who wanted to learn, but was facing difficulties doing so. I am considered to be a student with a permanent disability within our system, which for me means that my PTSD is considered to fall under “any impairment, including a physical, mental, intellectual, cognitive, learning, communication or sensory impairment or a functional limitation” that is considered to be “lifelong,” rather than needing to be re-confirmed every 12 months or so.

I was often granted the opportunity to either have an alternate format for an assignment, such as being allowed to view the distressing materials outside of the classroom, in an environment where I would be able to pause/take breaks as needed, or schedule a decompression session with a trained professional shortly afterwards, or tend to self care immediately after, such as taking a shower for grounding purposes- or there would be alternate reading assignments presented that still covered the same purpose of the evaluation, which was typically aimed more towards practicing synthesizing research and defending a thesis statement, rather than the exact topic. We had private rooms on campus where you could do so near the medical center and student groups area, or you could be excused to go home.

In the event that wasn’t possible, my professors would let students know both in the syllabus at the beginning of term that we would be covering potentially distressing materials (such as the impact of child sexual abuse in the event of psychology courses I took) in approximately the weeks it would come up, with a reminder before the pertinent class, and students were encouraged to take a breather in the hallway if they needed, or to not attend and to reach out to the professor instead to make alternate arrangements. If you found that you could not sit through the discussion unexpectedly, you were allowed to quietly excuse yourself and would not have that attendance held against you- in general, attendance policies are more lax than what I’ve heard from American universities.

This was because they recognized that some students face additional difficulties and extenuating circumstances when it came to their studies, and that working with a student to find an alternative arrangement was more equitable than forcing them to either fail a course or endure severe distress in the classroom environment.

There’s also a very big contextual difference in working through trauma and resilience building skills in a therapeutic setting, versus just randomly encountering distressing materials in your day to day life. Rapport between the clinician and the patient is critical for trauma work, (of which I have very personal experience between multiple professionals with varying specializations and therapy modalities). Having experience with exposure based therapies, and therapies that focus on desensitization and re-building neutral or positive associations: it’s extremely important that the patient feels like they are in control, that they have the ability to end the session if needed, and that they trust that the clinician has their best interest in mind and is there to de-escalate and co-regulate with the patient if the situation becomes too distressing. Sessions were often so exhausting I was told specifically to make arrangements to clear my schedule out for the remainder of the day and to focus on self care- specifically to get rest, and the maximum frequency of sessions I was regularly prescribed was twice weekly after three sessions a week was deemed too much to endure by my clinician.

Skills building in anticipation of severe physiological reactions, prior to informed, consensual, and skillfully monitored and administered sessions was a huge part of my therapy process. They don’t just throw you into the deep end unexpectedly, and it’s a process that takes time- in my case, years. A bad clinician, or unexpected exposure, can be re-traumatizing to patients, severely hampering progress. There is a lot of additional work outside of sessions patients do, especially in manualized therapies where the patient is given ‘homework’ assignments, and skills practicing so that they become more reflexive to reach for even outside of distressing situations. It took me over a decade to be capable of physically being inside of a car without it being intolerable, and even now, it’s very difficult.

Trigger warnings allow people the autonomy to choose what they are willing to engage, or not engage with. Additional arrangements can then be made to either allow them to endure interacting with the content, or they may choose not to, if it’s not something that they direly need to do, such as in a recreational context.

12 Likes
7

On the actual topic … trying to show the effectiveness or non-effectiveness of warnings isn’t useful because people who support warnings on principle can argue that the harmful warnings weren’t created properly or with enough consideration.

More broadly, I’ve said before that I support content warnings and trigger warnings but hold the unpopular opinion that I’d prefer audiences write them and publish them in their own indexes.

This is probably unacceptable to a lot of people … it’s admittedly selfish because I don’t want the pressure of writing the warnings and because I want to present my work as it is.

However, wanting other people to write warnings is also well-meant because I think that other people are in a better position to decide exactly what’s worth warning about. I hope @alyshkalia’s content warning wiki gains traction for that reason.

8

In addition to the points about context people have already made, I think framing is also worth discussing. In my own work, I tend to refer to these pieces of information as “content notes” rather than any sort of “warning,” both to try to offset anticipatory anxiety and because I frankly don’t want to presume how anyone is going to react to any given topic. The content in and of itself is not the danger, and as others have pointed out, trauma isn’t necessarily the only factor.

None of that makes the information contained in those notes any less potentially valuable to readers, whether they’re trying to avoid unintended, unskilled “exposure therapy” or they just would rather use their precious spare time another way. They’re not obligated to read my work, nor are they obligated to read the content notes before diving in if they prefer. Writing content notes takes me very little time, and including them as a visible but optional link on the title page feels like the most empowering option for the widest group of players.

(As a trauma survivor myself, I don’t tend to avoid potentially triggering works, but I appreciate having the option of not being blindsided - I confess I don’t always read CNs/CWs, and sometimes I’m fine and sometimes I regret it. None of this is empirical, but oh well, it’s what I’ve got.)

All this said, I do think audience-generated notes are also a good idea. I like what The StoryGraph does with it.

6 Likes
9

One of the positive effects trigger warnings are meant to achieve is choice - people can avoid things they do not want to experience at a given time and go towards the things they do. That’s a good principle to follow in general, and does not become bad simply because there’s a common term for it that helps people know what to expect.

Unfortunately the research pieces described for courses in academia, were at a low enough level that the material was likely to be considered mandatory by the students. At that point, trigger warnings aren’t being used to serve choice the way they are in most settings where trigger warnings are used. Academic courses tend to emphasise the areas where they want students to focus. Obviously indicating via a message that the focus should be on the triggering part of material (even if it turns out not the real aim of studying that book) is going to focus students’ minds in a way that simply reading a triggering book for purposes a reader can only guess would have that effect. A lot of modern non-discovery-based classroom-based education is built upon knowing that what one mentions tends to accrue more focus than what one doesn’t, and has done for decades.(For anyone wondering what a discovery-based classroom-based education is, it’s exploring content and allowing the lessons to either rise naturally or via careful questions that are adapted to where the student explores. If there is a true and meaningful choice of books, that weren’t being mediated by peer pressure, teacher/parent pressure or outright mandation, trigger warnings would be expected to be more helpful in that context).

Lab research exercises with rewards have a weaker version of the same issue, since by the time the content is revealed to be triggering, a commitment has already been made with a cost for pulling out (an effect that has been known since Zimmerman’s famous prisoner study in 1960).

The British Psychological Society (BPS) found in a meta-analytical survey from 2023 that the picture for trigger warnings more mixed:

  • People who are given trigger warnings are more likely to view the material than if the messages were not there, in studies where a choice was possible.

  • In light of this, the fact that people also averaged more anticipatory anxiety becomes more ambiguous. Are these people who actually fear the content more with trigger warnings, or are these people having the confidence to attempt content with trigger warnings that they’d simply avoid (thus falling off the research statistics, due to the way consent rules generally work in psychological research studies) otherwise?

  • The majority of people, even trauma survivors of the same material as triggered, who viewed material with trigger warnings weren’t triggered at all by the encountered triggers. This is believed to be because the process of triggering is more intricate in the majority of cases than whether it is there - context also matters.

  • For people who were triggered, the reaction is almost the same either way (some studies found it made a difference, both better and worse, but the majority found no statistically significant change). Trigger warnings, it turns out, neither reduce or augment the reaction itself. Neither do trigger warnings cause any change to people’s understanding of the material.

  • Limitations include significant definition problems (MPAA-style detailed warnings for films were excluded, despite these being arguably the most useful variety of trigger warning, yet the extremely vague “May contain upsetting content” was permitted), the remaining studies were so few, and so different in result, that any conclusions are tentative, and some questions about methodology.

For myself, my demo of Budacanta did not have any sort of in-game content warning, relying on other methods of describing and showing what to expect. I have a plan to have some sort of warning on the full game, with the ability to control some elements of it on-the-fly (because in practise a lot of people’s desire to voluntarily experience triggers fluctuates). Also, due to the history behind why that particular IF is being created, I’m hoping it will be played, at least sometimes, in settings where knowing there’s potentially triggering content will help people decide whether they are in a good venue that moment to play it without changing those settings. Those would be assessable questions by a research item, just not any of the ones I’ve encountered so far (including in this thread).

Third-party indexes are not something I can or should meaningfully control as an author, and I’m happy they exist because it is another sign that players (and readers and listeners) are taking control of their own experiences. I have been known to use TV Tropes to pre-screen recommendations for people who I know are triggered by certain things, even though it’s not formally designed for that purpose. Nonetheless, I think it would be best if I wrote something, especially given I’m planning to have in-game control of some of the triggers. Given the full game is hoped to be commercial, it would be reasonable for me to budget for someone to check the game for triggers after showstopping bugs are removed.

7 Likes
10

Is there a fundamental difference between a trigger warning and a content warning? I personally prefer the term “content warning” because I don’t like the assumption that because I’d prefer to avoid something that I am “triggered” in some negative way by it. Not that being triggered is bad or anything; I just don’t like people assuming that preferences are big emotional things.

@OverThinking 's recent game, Sundown, really affected me emotionally. I cried while playing because it hit me so close to home. I didn’t read the CW and later I wondered if I should have, or if I should start paying more attention to them (I generally don’t). But then I realized that it wasn’t a bad thing to play a game that made me cry and reminded me of a hard time in my life. I want to stress that this is me rambling about myself only and that I certainly don’t think everyone should play games that are upsetting to them.

As for studies: it can be true that studies generally show a trend, AND it can be true that some individuals find themselves outside of that trend. These are difficult things to study. Teasing out all the variables might be impossible. Disclosing sexual content, or disparagement of a religious belief, or child abuse, is a favor to readers, and not just because they may be “triggered.” It’s because they just might not want to.

11 Likes
11

I think that’s brilliant.

12

I think it probably doesn’t help that the term ‘trigger’ has become so loaded in recent years, particularly when it’s now commonly used in certain sections of society to deride and mock people.

From as far back as I can remember, every film I’ve ever watched or traditionally-published game I’ve played has had a content warning of some sort on the box - violence, sexual content, etc. - so I don’t see the issue at all with the same principles being applied to DIY/indie titles.

I think there’s so much to be said for freedom of choice - both the choice whether to view the content warnings or not, and the choice to engage in a piece of media or not once people are aware of what it contains. Everyone gets to choose how they interact with it.

7 Likes
13

I really understand people advocating that mandatory CWs are a burden or an obstacle or might be an excess. But I still think they are useful. As I said elsewhere, the option to see the CW or not is a nice method to satisfy both groups. It doesn’t force the CW onto the audience, but it helps a lot of people.

2 Likes
14

Sometimes in delivery, no. One of these studies I shared even shares the two.

Semantically, the difference has become night and day, which is what has made the two things different things.

It was the drawing of the medical term Trigger from PTSD to apply to any emotional thing that might temporarily disturb someone, non PTSD-patients, that set trigger warnings apart. People who aren’t clinical PTSD patients began to ask for trigger warnings for anything. The culture of content warnings has been centrist, of the kind @pbparjeter described.

I asked people to read a ton and some did. I don’t want to take much more time or space. But I feel I should probably clarify where I stand personally, which is that I am for content warnings. I am against trigger warnings.

I believe that expectation of the warnings, and use of the word Trigger are countertherapeutic. A trigger fires the bullet automatically. The whole point of cognitive behavioural therapy, which is empirically established, is that there is a point between the trigger and the bullet where you get to do something with the signal with your mind. Trigger warning culture promotes avoidance of that, and asks other people to control their behaviour instead of you engaging with that inbetween point in your emotions. It also has a contagious quality in that it asks others to respect and share the same countertherapeutic idea. It is entirely at this fundament level that my opposition grows.

So in case what I said in the above paragraph isn’t clear (my fundamental opposition idea, rather than extended contxtual ideas re: IF) the context is life, not just IF. I’m all for putting content warnings on IF.

Honestly, this is a problem. Research leads us towards certain kinds of truth, or just the truth, but we’ll choose to ignore it where people we know are concerned. People enabled my social anxiety, my family’s alcoholims or drug use, etc. Enabling avoidance of problems is harmful. That’s the bottom line of trigger warnings for me. It’s a fairly elaborate culture constructed on avoidance of psychological engagement. It also asks other people to catalogue the bumps of the world particular to individuals they don’t even know. The list of dangers is endless, the specificity of people’s problems endless. This is why the centrist model is important.

Anyway, thanks to anyone who read the studies, teased out their details and thought about the whole situation. For me it’s important to say this stuff and I’m glad I did, but I don’t want my life or yours on this forum to be an unending argument about non-IF-centred stuff (My best friend is a debating coach and it could be argued his is.) so I won’t post to this topic again.

-Wade

8 Likes
15

Wade has mentioned he would not post again to this thread, so this isn’t seeking an argument: merely to note that there is growing evidence that unmodified CBT, (cognitive behavioural therapy) despite historic application, may not be a good fit for C-PTSD patients, and several of the psychologists and psychiatrists I saw who specialized in treating PTSD (and would have specialized in C-PTSD, save for that distinction is not legally recognized in Canada due to our diagnostic handbook, though there has been discussion about its inclusion in a newer addition to more closely align with international diagnostic tools) would not use CBT with a patient who had a diagnosis of PTSD but dually would qualify for C-PTSD.

The difference between CPTSD and PTSD largely is that C-PTSD is thought to come from chronic exposure to trauma, rather than a distinct event, as with PTSD- a common origin would be an abusive childhood, though it can also arise from say, domestic violence- but when paired with the developmental trauma from a childhood onset, greatly complicates the course of treatment, due to core beliefs and extensive maladaptive coping mechanisms that are common in patient presentation.

There are modified modalities that fit under CBT that have far more efficacious outcomes with traumatized populations: DBT and CPT were the most commonly used modalities among the professionals I saw. Unmodified CBT in of itself was viewed as something that non-trauma specialists would often default to, as it can be very effective in treating anxiety disorders (of which PTSD is classified as one), (and is easy for non-specialists to offer to the general population and for insurance to bill for due to its regimented structure), but in their own practice, saw that it was not as effective as more targeted modalities that were trauma focused or trauma informed therapies.

DBT, or dialectical behavioural therapy, was drawn from to manage the physiological symptoms and maintain a larger threshold of tolerance- to self-regulate distress, and work on challenging underlying maladaptive schemas: though this modality is more commonly used in BPD, or borderline personality disorder patients, it’s generally seen as very effective when used in complement with other modalities because of the physical symptoms of C-/PTSD often interrupting or requiring the cessation of a session.

CPT, or cognitive processing therapy, was largely about being able to form a coherent timeline and narrative to an underlying trauma, and challenge the deeply ingrained beliefs that surrounded the traumatic event and its influence on the patient, to restore a more balanced perspective.

I don’t have experience with TF-CBT personally, (trauma focused cognitive behavioural therapy, different from unmodified CBT), but I believe it has similarly positive outcomes for those with complex trauma histories, as opposed to plain CBT, which is often additionally pushed because it has been studied for a larger amount of time than other modalities (and thus more providers are also trained in it.) PE (prolonged exposure) is also a modality that I’ve heard positive outcomes from, and EMDR (eye movement desensitization and reprocessing).

One of the biggest issues with the existent research is that C-PTSD and PTSD patients are not distinguished between in the diagnostic criteria (in at least the American and Canadian context,) despite the different origin of the conditions and varying manifestations, making it difficult to assess what may better meet their needs.

4 Likes
16

Wade, I know you said you wouldn’t post anymore here and I don’t expect you to, and I’m going to try to say my piece here without giving you anything you feel pushed to respond to. I generally agree with your take and I think this is an important and interesting conversation to have. But I’d like to give everyone a gentle reminder that this kind of statement is a surefire way to get someone to not listen to your point. We all have places where we are entrenched in our viewpoints and experiences. I try very hard to have data-driven opinions, which can be difficult when data is constantly changing (I taught Biology for years and a lot of stuff I taught back in the late 90s turns out to be wrong or incomplete, like the ill-fated “central dogma” of molecular biology, which turned out to be-- like most dogma-- incorrect). But I still smoke cigarettes sometimes. I drink too much. I sometimes give too much weight to my own anecdotal experience and have a hard time squaring it with established fact. I sometimes behave in ways that are wildly contrary to every data point I know to be true about how to live and think. So does everyone. And I’m talking about things that are without a doubt factual, like the fact that cigarettes will fucking kill me. Most social sciences live in more gray areas, and this is still a gray area. Again, I think you’re probably right, but I don’t think we can call the game there like we can call the game on tobacco.

I spent a lot of years arguing with people about evolution and reproductive rights and the poison of the patriarchy, and I have had all the research on my side every time. And it has never, ever done a single thing except make people dislike me and be less likely to listen. People believe what they believe, and they have their reasons, and at my advanced age I have come to understand that telling them they’re a problem for that will close minds to any point you’d like them to consider. I say all this with the utmost respect for both you and Tabitha.

8 Likes
17

That’s twisting his words.

2 Likes
18

I thought about this some in light of @AmandaB’s comment on this being a grey area. I realized my previous comment might seem one-sided.

I wrote:

… Trying to show the effectiveness or non-effectiveness of warnings isn’t useful because people who support warnings on principle can argue that the harmful warnings weren’t created properly or with enough consideration.

The way I put it, it sounds like I’m singling out people who support trigger warnings by suggesting they use some sort of faulty logic to justify what they want.

I wanted to add that people who are opposed to warnings on principle can also discount the question of effectiveness in a similarly questionable way. Such as by arguing: “If we found that people are benefiting from warnings, they just haven’t seen how badly they can be applied, and that makes it worth abolishing the system entirely.”

I think either sort of argument falls under the category of “begging the question” or assuming the conclusion. IMO, whether there is evidence for a position is always secondary to this sort of argument. It’s a fallacy, and it can be used to justify extremes.

At the same time, it’s also a way of expressing a principle or ideal rhetorically, and it’s a way of suggesting that the factual situation painted by the evidence can be overcome. So I think both sides can have a point regardless of the evidence they present.

19

I mean, I’m not a psychologist or a therapist, but I have no issue with calling them “content warnings” instead of “trigger warnings”. If the issue is specifically with the term “trigger warning”, that seems like a trivial problem to overcome.

7 Likes
20

I took it as Amanda speaking about herself, and how expounding on her beliefs re: evolution, reproductive rights, and patriarchy often lead to the insinuation that those she was speaking to were the problem and not just their beliefs.

And you’re right; he didn’t say that. I took Amanda’s post as a cautionary statement to suggest he should avoid straying that direction rather than suggesting that’s what he said.

I sincerely detect no hostile intent or subtext in her post.

That said, I can be a socially stunted dolt at times, so who knows.

5 Likes