Don't Feed the Trolls

[Aviastroitel]

There just needs to be higher standards.

Isak Heartstone, Thomas Dambo's Breckenridge (Colorado) Troll. https://trollmap.com/ lists ~ 50 other trolls the Dane has created around the world.

 

[Kelliezgirl]

Thanks for sharing the article. I’m just pasting the citation and abstract below for info - it’s linked from the article but I’m not sure that everyone commenting in the thread has found it yet.

Brown, M. A., & Toyama, M. (2025). Inside the link between narcissism and social media trolling: the involvement of malicious envy and exposure to antisocial media content. Behaviour & Information Technology, 1–14. https://doi.org/10.1080/0144929X.2025.2547922

ABSTRACT​

Considering negative psychological and other consequences of trolling on social media, the present study addressed the link between narcissism and social media trolling and its potential mediators, including malicious and benign envy and exposure to antisocial media content. Participants were recruited via an online crowdsourcing platform, Prolific, and 326 (aged 39.4 [SD = 12.4]; 48% male, 50% female, 2% non-binary or non-disclosed gender) completed our online questionnaire without attention-check failures. Our mediation analyses (controlling for covariates) indicated that malicious envy mediated the links of narcissism to social media trolling and to exposure to antisocial media content. In addition, exposure to antisocial media content also served as an additional mediator, as narcissism was linked to social media trolling indirectly through malicious envy and then exposure to antisocial media content. Narcissism predicted social media trolling directly, even after additionally controlling for the two mediators that were also directly linked to the trolling outcome. These findings provide practical implications, highlighting the importance of addressing the identified antecedents of social media trolling, and indicate future research directions to investigate additional associated factors and explore and develop approaches and interventions to reduce trolling on social media.

Thanks for providing this. One of those moments I'm glad I went to a quality University and still have journal access.
I thought this part was interesting.

2.2.5. Social media trolling​

To measure social media trolling, we adopted the nine-item Global Assessment of Facebook Trolling (GAFT) (Craker and March Citation2016), while replacing references to ‘Facebook’ with a general term referring to ‘social media’. Using a Likert scale, participants indicated their level of agreement, ranging from 1 (‘strongly disagree’) to 5 (‘strongly agree’), for each item (e.g. ‘Some of my funny social media posts could be considered offensive by others’). Three items (e.g. ‘I prefer not to cause controversy or stir up trouble’) were reverse-scored, and then all items were summed for a total score, with higher scores indicating higher levels of social media trolling. The participants’ scores ranged from 9 to 43, and their mean score was 19.1 (SD = 6.4). Cronbach’s alpha for this scale was .77.

 

[ShelbyDawn57]

However we do it, we should steer clear of painting it as any derailment of a conversation or thread.

Because if we went with that, then this forum would consist of nothing but trolls

1. What is your name?
2. What is your quest?
3. What is the air speed velocity of an unladen sparrow?

 

[Kelliezgirl]

Yes. There are authors who proudly tell us they're writing stories to deliberately stir the outrage, mostly in the Loving Wives category. One of many reasons to do some homework before you post content there.

Seems like a pointless exercise to me, like shooting fish in a barrel, but some folk seem get their jollies from it. It's odd behaviour.

Agreed.

 

[Kelliezgirl]

1. What is your name?
2. What is your quest?
3. What is the air speed velocity of an unladen sparrow?

And we're off...

 

[WCSGarland]

1. What is your name?
2. What is your quest?
3. What is the air speed velocity of an unladen sparrow?

African or European?

 

[SamanthaBehgs]

Yes. There are authors who proudly tell us they're writing stories to deliberately stir the outrage, mostly in the Loving Wives category. One of many reasons to do some homework before you post content there.

Seems like a pointless exercise to me, like shooting fish in a barrel, but some folk seem get their jollies from it. It's odd behaviour.

I thought we had a category for this under Humor/Satire. Guess I was wrong there.

 

[SamanthaBehgs]

I'm not sure how you "feel my frustration" when I'm not actually frustrated.
I find the type of behavior that results in people running and hiding while screaming "troll" amusing, and to be honest kind of sad.

I'm remarking on the inevitable consequences of posting an article like this one here. I think that has some relevance to the topic at hand.

Digging into it a little deeper, it's difficult to try and psycho-analyze people via questionnaires, particularly when it's a self sorting process.
I've got a couple of friends who consider themselves serious shit posters, and they'd probably have a field day gaming the system if the opportunity to take a survey like that crossed their paths.

Another instance of an important correction. What I read in your posts as frustration is actually amusement and...what emotion would you call if if you find the behavior of others sad?

I don't agree that the consequence is inevitable, but that's not really something we can prove or disprove. The logic would be circular. If it happens, you can call it inevitable, that it would have always happened that way. But just because it happens doesn't mean it has to have happened that way. So we're stuck with another question but I'm not certain how we'd answer this one.

I am, however, having a bit of trouble between two things. In this post, it seems to me like you are saying that you cannot conduct psychological research that involves any sort of self report (that self sorting process of questionnaires). That would invalidate much of the research being done in the field, but I know from previous conversations with you that you value looking at the research in a given field. Are you expressing a disdain with the majority of research in the field of psychology? If we are rejecting any research that involves self-report, then none of this (or anything below) is interesting in the slightest and, honestly, you and I should not discuss research in the field of psychology because we disagree on something so fundamental as to make the discussion pointless. This is not really the forum in which to debate that. I'm okay with accepting this if you are if that's the argument being made.

Thanks for providing this. One of those moments I'm glad I went to a quality University and still have journal access.
I thought this part was interesting.

2.2.5. Social media trolling​

To measure social media trolling, we adopted the nine-item Global Assessment of Facebook Trolling (GAFT) (Craker and March Citation2016), while replacing references to ‘Facebook’ with a general term referring to ‘social media’. Using a Likert scale, participants indicated their level of agreement, ranging from 1 (‘strongly disagree’) to 5 (‘strongly agree’), for each item (e.g. ‘Some of my funny social media posts could be considered offensive by others’). Three items (e.g. ‘I prefer not to cause controversy or stir up trouble’) were reverse-scored, and then all items were summed for a total score, with higher scores indicating higher levels of social media trolling. The participants’ scores ranged from 9 to 43, and their mean score was 19.1 (SD = 6.4). Cronbach’s alpha for this scale was .77.

Assuming you aren't, then I have some thoughts going further. In this post, we are looking at the particular definition of social media trolling and how it was defined in the study, using a validated measure of trolling that was adapted to include all of social media instead of simply Facebook and had good internal consistency. Is your second post here designed to indicate that you cannot possibly use this measure in any meaningful way? I wholeheartedly disagree with this - as you cannot simply say the data aren't the data simply because you don't feel the researchers have identified what is being measured, especially when the assessment is internally consistent. Clearly *something* is being measured here, even if it's not trolling.

Let's assume, for a second, that this is not measuring social media trolling and is instead measuring some other concept. That is one thing that I think I am reading in your post, though I admit I could be wrong (and have been proven wrong on other things regarding reading into the subtexts of your posts, so I welcome correction here). But assuming it is the case, or even just assuming it's an interesting question, what would we call this concept if it is not social media trolling? Because we are still seeing a relationship between narcissism and whatever we decide to call the concept measured by this questionnaire and identifying malicious envy as a mediator (and further documenting a multi-step process in this model that goes into media consumption) while continuing to show a unique relationship even when these important factors were controlled for. In other words, they admit they don't have the full picture. And if this isn't social media trolling, knowing what this concept is would be a necessary piece of better understanding this picture. In short, it invites curiosity either way and, to me at least, is interesting. Perhaps it's not to anyone else here, in which case, that's fine. I am glad to let it rest and continue to find it interesting on my own. It won't be the first time nor will it be the last.

Either way, I'm also now craving parfait so I think I'm going to go see if I have the fixins.

 

[dream_yahoozie]

Twisting Reality Online LoLS

That seems a little accurate

 

[Kelliezgirl]

There is a pretty significant replication crisis in science these days, and the social sciences are amongst the worst.

The data they gathered was from the website Prolific, which pays people to take the surveys and so forth.

https://www.prolific.com/


I think it's reasonable to question to what extent the type of people who participate in that are representative of the population as a whole.
For example, they are very vocal about their use in providing data for AI. I suspect that means the vast majority of the AH would nope out of working with them.

Does that mean I think we should disregard the study? Of course not, but I do think we should approach it with caution.

Definitions are critically important in any type of research. If you don't have a valid definition, and it doesn't accurately reflect what you are attempting to measure then the whole exer is for naught.

In this case, using a very loose definition of trolling reasonably calls into question the validity of the results.
Does making a joke that SOME people MIGHT find offensive really equate to trolling?

 

[SamanthaBehgs]

There is a pretty significant replication crisis in science these days, and the social sciences are amongst the worst.

The data they gathered was from the website Prolific, which pays people to take the surveys and so forth.

https://www.prolific.com/


I think it's reasonable to question to what extent the type of people who participate in that are representative of the population as a whole.
For example, they are very vocal about their use in providing data for AI. I suspect that means the vast majority of the AH would nope out of working with them.

Does that mean I think we should disregard the study? Of course not, but I do think we should approach it with caution.

Definitions are critically important in any type of research. If you don't have a valid definition, and it doesn't accurately reflect what you are attempting to measure then the whole exer is for naught.

In this case, using a very loose definition of trolling reasonably calls into question the validity of the results.
Does making a joke that SOME people MIGHT find offensive really equate to trolling?

I'm familiar with the replication crisis, having taught about it but also about the fascinating questions it brings up. For example, why are old studies not replicating today? Is it because they were never valid to begin with or because there is some interesting difference that we can dive into? For example, there was a study on stereotype threat conducted in a non-diverse area that failed to replicate at a very diverse institution many years later - for many it was the poster child of the replication crisis. Once someone really got into it, though, the problem wasn't that stereotype threat wasn't a thing at all, but that the context in which we measure it matters. It only affects performance in certain settings.

I think the same is true here. And prolific is a new enough site that you may end up being right about it, but I don't think throwing out the entire study because of its use would be the best way to handle it. Otherwise, we'd have missed out on some valuable research using Amazon's Mechanical Turk model (which turned out to be validated against later clinical interviews of those using the platform) or some of the really cool "email to your smart device" research that was originally decried because it wasn't done in front of a researcher/in a lab. It's another interesting thing to note, but doesn't in my mind completely invalidate the work. And given that all research needs to acknowledge the limitations in their sample in one form or another (so many studies are based on W.E.I.R.D. people and need to be replicated in more diverse areas at some point), that's kinda the point of replication. The "replication crisis" is such a big deal not because it means nothing is true, but because it means we need to start publishing more replications so that we can encourage more replications. Cross-cultural and cross-method research has, largely, been immune to this lack of publication bias, however - journals have always found it interesting to publish these types of replications.

You and I agree, however, about approaching the study with caution. But that is true of all research, regardless of the field of study.

And part of that caution is absolutely about what you find to be a loose definition of trolling. That would call into question what concept they are measuring, to be sure, and I will be honest when I say I have not dived into the validation studies done on that particular questionnaire, though it is cited in the methods section you pasted here. However, if we grant that it's not measuring trolling for a momen, it's still measuring *something* and that's interesting. Andn that something is related to what and how we behave when we post online, also still interesting and relevant for an online forum. Does it derail this thread a little to talk about what that something might be? Absolutely. But it's my thread, so I'm happy to derail it. But if that's not interesting for you to talk about, then, again, it's something I'll ponder on my own. Or if a group of folks from the AH ask that it be dropped, I'll do that. It's my thread but not my forum, and I certainly am not here to annoy people.

Ultimately, however, the definition, though important, does not invalidate the data. Let's say I misdefine depression as sloth but still use the BDI-II (a well validated and reliable measure of depression that, although not used diagnostically is a strong predictor of diagnosis). Then I talk about how sloth is related to increased cortisol levels and decreased ability to fight off infections. Have I misdefined a variable? Certainly. But there is still this thing that is measured by that measure that has this effect. The data are still the data. Numbers don't change just because I've interpreted them or labelled them in a way you wouldn't. It just means we have an interesting and valuable discussion about how we *would* label the variable.

Factor analysis is a GREAT example of this. The factor structure forms mathematically, but what you call each of the clusters (I got sick of saying factor) is a matter of interpretation, and one that is interesting to discuss. At least to me. If it isn't to you, that's fine. I won't be insulted and can ponder it regardless.

PS: not sure what happened for it to quote you twice. Sorry about that.