 Hello everyone, it's good to see you here today, that even though there was a bit of snow today that you still made it and came all the way here, that's really great. My name is Hannah van Den Bosch, I'm a program maker and creator for Studium Generale. We organize all kinds of events like this one on events and on topics that matter. So now to today's topic, online haters and why internet discussions escalate. I think maybe you know this when you go to a comment section of a YouTube video or on Twitter or maybe a news article. It can get a bit rough there in the comment section. Sometimes you see a lot of fighting, lots of swearing as well, but why is that? Why do discussions online tend to escalate more quickly than when you meet someone in person? It's an interesting question I think. Someone who can tell us more about this is Carla Rose. She is an assistant professor at the Department of Communication Rights here at Tilburg University and she has a PhD in social psychology. Her other research and expertise is focused on online discussions and also the behavior in interaction partners. So please give a big applause for Carla Rose. So yes I will just start by saying that this topic of online discussions was the topic of my PhD research which I conducted in Groningen in the north of the Netherlands and which I recently completed in September so it's quite fresh research and I'm currently also still working on developing this further so if you're interested let me know to maybe discuss further and today I will talk to you about why do online discussions tend to escalate more easily than face-to-face discussions and my talk will be quite scientific because I understood that you are mostly students here at the university or somewhere else at the university so also when something is not clear please let me know. You can also just raise your hand during my talk and I can try to elaborate a little bit more or explain something when something is unclear so feel free to also raise your hand during the presentation. So let's start so as we've already said multiple times now we see that discussions seem to be more prone to escalate in misunderstanding and conflict when they are conducted via online media and mostly textual media as compared to when people discuss face-to-face and this raises two interrelated questions so first of all why does this happen? What is causing this online escalation and secondly building on this when we learn what can explain this online escalation maybe we can also learn how to solve it or prevent it and according to these two questions my talk will also be separated into blocks so the first part will be about why does online escalate so our research about that and secondly how we can maybe prevent or reduce this online escalation but first I want to learn your ideas about why online discussions escalate. No direct consequences okay people can wear a mask and are able to share their frustrations with less risk of long-lasting tension so that's probably what this anonymity is causing so no accountability consequence of your actions are not direct you don't see the reaction of your comment to the other person so maybe it's less hurtful problems on others yeah okay so I see a lot of yeah not real context anonymity people not feeling accountable people not maybe feeling emotional connections maybe feeling that they can do whatever they want because they are anonymous so thank you for this input so I think I will just continue to put the rest of the talk so thank you for your input so this is nice because it also ties into one of the common explanations that we often hear for online escalation I have two so I will start with the other one oh no I will start with this one yeah that's true so this is tying into what you also mentioned so it's often thought that because people are often anonymous in many online discussion places that they feel that they can say whatever they want without any consequences for their impression management so they feel they can start to call names or say offensive language without any consequences and this is formalized in the online disinhibition effect so it's also often written about in a literature and second of all it is often assumed that people when they are expressing themselves in text are less clear because they cannot use their body language like gesturing or smiling so no facial expressions as well and this will then lead to more misunderstandings and this can maybe escalate into conflict and this is also visible in a literature so we hear it people in popular opinions but also in a literature in the media richness theory but then yeah when we look at this these sort of explanations and then literature more broadly we can summarize this as as follows so it is often assumed that the medium characteristics so text-based for example will change people's psychology and therefore this will change their behavior so in the example of the disinhibition effect so it is often assumed that when people are anonymous they become less concerned with being positively evaluated by others or be less concerned with others in general and therefore they will start to act differently so they will do whatever they want whatever they feel like and this can result in this inhibited behavior and maybe take on a toxic form when there's this controversial discussion going on so often this line of reasoning is assumed but we propose that maybe it can also be the other way around so we assume that when people go online they might not necessarily change psychologically speaking and like maybe they don't become disinhibited but we thought that what does change when you are communicating online is also your behavior so people's behavior is limited maybe online and this can have social psychological consequences and I will explain more about this later but we first need to know we first need to look at behavior because we saw that discussions when they are held online tend to escalate more easily than when they are held offline right so what might be informative is to look at what people do actually in offline conversations to see how they there manage their disagreements without escalation because when we see behaviors that people do offline we can also maybe think that of see that certain of those behaviors are not possible online which might then explain this escalation so how do people deal with disagreements face to face and there's two important behaviors distinguished here so first of all responsiveness so this is defined as agreed to which interaction partners provide each other with instant feedback so for example you might notice it yourself as well when you are having a discussion with somebody and this you want to show that you are listening to this person you start to nod or say mm-hmm or yes so you're responding to the other person in essence and in this way you show that you are listening and engaged in a conversation and also interested in what they want to say and maybe also show some kind of understanding that you see what where they are going and secondly what people often do when they have a controversial discussion they tend to express this their disagreements in a very tentative manner so they instead of being very direct people start to use a lot of hesitations hedges and qualifiers so they will say like I sort of agree with you but I also not totally agree with you maybe because on the other hand right and like very fake just a lot of words without much content so in this way they show that they are maybe still a bit uncertain about their opinion and can also still maybe be convinced of another opinion which leaves room for different opinions to be brought up in a discussion and secondly they also show it is ambiguity and tentativeness that they don't want to offend the other person so they want to maintain good social relationships and because this ambiguity might still be a little bit ambiguous to you I also included an example of this so this is an example of a real face-to-face comment made by one of the participants in one of our studies so the opinion is not that clear and I also remarked all the expressions that I saw that indicated this yes this doubt and ambiguity so not really a little bit so qualifying a lot and also saying a lot maybe and also and say an um so all those little words and expressions that show this ambiguity and people do this really a lot so when you when I listen to the recordings of the face-to-face discussions it seems very natural but when you write it down on text you see just how much people do it and then it seems quite strange actually um but yeah it has a function so when we come back to those two behaviors is responsiveness and ambiguity we termed them diplomatic behaviors because these behaviors enable people to have a discussion about a difficult topic on which they disagree maybe without damaging their relationship so they can be diplomatic in this way and then this makes us wonder because when these behaviors are so important to maintain this good discussion face-to-face maybe people are limited in these behaviors online and this can then explain why online might escalate so that is what we proposed not very coincidentally um so we proposed that um now again you see the model that we proposed right so first the medium changes behavior which changes psychology so we proposed that um two of the characteristics that differentiate online from face-to-face discussions will limit people's ability to be diplomatic to be responsive and to be ambiguous so first of all many online media are asynchronous so responses occur with long delays or people are maybe also entering messages at the same time by which it becomes like cross talking a little bit um and this then reduces responsiveness and second of all a lot of online media are text-based so people are expressing their opinions only in text and this will probably make the expression less ambiguous people might be more likely to say I disagree rather than this slight filling example that we just saw that's that's a lot of typing work that people don't do that and then because responsiveness and ambiguity are so important in face-to-face discussions to maintain this sense of agreement and good social relationships we thought that this lack of responsiveness and lack of ambiguity in online discussions will explain why people experience less agreement and less harmony so a worse relationship in essence and then as a sort of an underlying process for this we thought this might be related to experiences of feeling ignored in the online conversation because your partners are not responding to you and second of all also because maybe people think that others are disinhibited and not being concerned with them because they are not responding to them and also expressing themselves very clearly which might come across as very opinionated and not very socially concerned so this is what we expected and we tested this in a couple of studies two of which had this how do we say design so here we had students come into the lab in small groups of three or four and discuss about controversial topics like the refugee crisis or so and they were unacquainted so they didn't know each other before which is quite important when you want to assess relationships evaluation etc and they had discussions of both via text-based online chats and face-to-face and in the second study we also had a confederate come into the lab an actress and that was also participating in this discussion without the participants knowing so there were just one of them was actually with me and she was pretending to be a quite a right-wing student with a quite a strong opinion so and the other students were mostly left-wing so we tried to get a little bit more of yeah controversy and disagreement in this sense and I will also present the results of the second study shortly but in both conversations we then or in both studies we then looked very closely at behavior so what do people do and how do they express themselves and we also asked them about their experiences of this conversation and about their affiliations of each other in questionnaires yeah so onto the results first of all for for the behavior the first look we were interested in is that we indeed saw in line with what we expected that people were less responsive in the chat conversations and in the face-to-face conversations so there was more of this cross-talking and people were not responding to each other that much as they did face-to-face and we also saw that the online chat messages were way less ambiguous face-to-face messages so people express themselves with way less hesitations and hedges and qualifiers than in the face-to-face comments and what is interesting as well is that we didn't encounter any sign of disinhibited behavior so we didn't see people calling names or being aggressive which is maybe not very surprising because it were just students having a conversation lab for study points or so so they were probably motivated to keep a nice conversation but indeed we saw no disinhibited behavior so in line with what we expected we see that indeed online messages are less responsive so people in this discussion are less responsive to each other and also there's less ambiguity in their messages so how will people experience this behavior was the second question so what is the psychological change so here we saw that when they engaged in a chat conversation people felt more ignored than in a face-to-face conversation and also that they perceived each other as disinhibited more disinhibited than in the face-to-face conversations and this last thing is extra interesting because we just saw that most yeah so there was no approval of this inhibited behavior in the content of the conversations but still people thought their partners were more disinhibited than in the face-to-face conversations and also they didn't consider themselves to be disinhibited online so this is quite interesting apparently something happens within these discussions that makes people think that their partners are disinhibited while they are actually not or while they think they are not and while while they are also not behaving in this way so that is also supporting our expectations that people will feel more ignored in these online discussions than in the face-to-face discussions and also perceive each other as disinhibited and lastly we also looked at how this affects people's sense of agreement and social relationships so harmony and we also saw that people felt less agreement in the chats than in a face-to-face conversation also less harmony so they experienced less identification between each other after the chat conversations and for the agreement it's actually interesting that we also looked at when we looked at the actual opinions of interaction partners we didn't see that they were more disagreeing with each other so when we measured their private opinions on the discussion statement these were not more different than in the face-to-face discussion so there was not actually more disagreement in the chats but still people perceived more disagreement in the chats than in the face-to-face conversations but in general we find support for our model that there's less agreement and less harmony experience online so and yeah because we expected that these behaviors this less responsiveness and less ambiguity will yeah lead to or be related to this perceived ignoring and disinhibition etc we also looked at the correlations between these concepts and these were in line with expectations in that we just see that people experience so while they do not actually disagree more they do experience more disagreement in the chat conversations because they express themselves less responsiveness less responsively and less ambiguously and this less responsiveness and less ambiguity leads to feeling ignored and disinhibited so there's in a way these behavior so this lack of these diplomatic behavior seems to lead to all kinds of misperceptions and misattributions so people while they are not this inhibited they think that their partners are disinhibited because they are just expressing themselves so clearly and not very responsiveness responsively and also while they are not actually disagreeing more they experience this disagreement because they are having this discussion in which everybody is expressing themselves so clearly sorry opinions so clearly and so little responsively so in in effect we see that the medium actually changes behavior but people seem to attribute this behavior change to each other's intentions instead and that's going to have consequences so this brings me to the the conclusion slide of the first part so the first question was why does online seem to escalate more easily than offline conversations so first of all we saw that it's often expected that it is due to online messages being ambiguous or fake unclear and and or that this was due to people being disinhibited online because they are anonymous anyway but we found that also the opposite can be the case so we we saw that when we look actually at behavior we see that online comments are actually more clear than face-to-face comments and that this clarity can lead to misperceptions of disinhibition in others and and make people think they disagree more so we see that the medium seems to make people less able to be diplomatic but that people misattribute this lack of diplomacy to each others yeah sort of asocial motivations so is this clear until now so what do you think makes the difference whether the only expression is too clear or too vague do you think it's it depends on the context of the discussion for example when a discussion is about for example refugee crisis do you think when people express themselves too clearly instead of getting into these ambiguous expressions that's that's more enhanced face-to-face do you think it's easier in this context easier in the online context or easier in the context of the topic or in the online context yeah now so we saw that or yeah I think and we also saw that in the online discussions it's more difficult to be ambiguous you could do it you could also type um and maybe and or so but yeah it's not just not what people do so in that sense it's more difficult do you understand or is it asking your question as well or my question was actually about the context okay yes for for example or for example covid yeah whether covid is real or when whether people are too clear in for example uh stating data from previous research from previous research about covid yeah and other people still doubting yeah yeah that's a very good point the context in general is of course very important here because where we are looking at here uh is uh at controversial topics discussions about controversial topics because when you um so in that in that sense then ambiguity is nice because then it leaves more room to have different opinions uh and disagree with each other in a harmonious way but when you are having a discussion or a conversation about something about which you completely agree then it would be preferable to say I completely agree in a very clear manner because when you would then say I maybe sort of agree then it makes it probably less convincing that you actually agree so it really depends on what you're discussing and here we are looking at yeah controversial topics so in that way it's really important indeed the context yeah thank you so um but yeah we looked at why online escalation happens but then this can bring us to uh how we might be able to maybe prevent or reduce online escalation that would be nice right so maybe we can use all this knowledge that we gathered until now and try to maybe uh make online a better place so we thought based on our research to try to increase these diplomatic behaviors in the online conversation so maybe this will help to make them more harmonious so maybe we can we can find a way to increase crease ambiguity and increase responsiveness in textual chats and this we did in a quite of a creative way by transcribing face-to-face and online discussions that we collected in the previous study we took these conversations and we put them all in the format of a textual chat so we had half of the conversations were actually were already already original textual chats so we remained them in the same manner but we also took the face-to-face conversations and put them in the format of a textual chat so we transcribed them and also gave usernames so as you can see here as well so this is actually an original face-to-face conversation that that we made appear as if it is an online chat conversation and we did this in order to try to increase the ambiguity and increase the responsiveness because we saw in the face-to-face conversations that there is more responsiveness and ambiguity so we yeah in a way we inserted ambiguity and responsiveness in these textual chats but we also had the original textual chats to compare right and then we had a group of uninformed observers to evaluate these discussions so the online original online discussions and the original face-to-face discussions and these people thought these were all original chat discussions all original online conversations and we asked them to say to see to what extent they saw these diplomatic behaviors to what extent they thought these were responsive and ambiguous and we also asked them to what extent they thought these people in these conversations were yeah having a good relationship and being in agreement with this order and that gave the following results so in line with what we would expect based on the previous we saw that the more responsive these observers thought that the chatters were the more they thought these chatters were having agreement being in agreement and having good social relationships so that's in line with what we would expect but contrary to expectations so completely other way around was the effect of ambiguity so we saw that the more ambiguous the observers thought that the chatters were the less agreement and the less social harmony they observed there to be in this chat context so this complete this is the complete opposite of what we found before before we saw in the face-to-face context that ambiguity helped to make the conversation more harmonious and agreeing but in this chat context it seems that maybe this ambiguity is perceived as a bit yeah weird or unfitting and maybe when we think back to this previous slide with this long comment that was original face-to-face you can also maybe recognize why it might not have worked because this ambiguity can also just seem very unnatural in this textual format so maybe that was going on so we can see or conclude that so responsiveness might work but this face-to-face ambiguity might be a bit weird online and that makes me thought for yeah maybe we can find another way to make online more ambiguous in a way that is more fitting to the online medium to the textual medium so instead of inserting this face-to-face ambiguity with um and maybe and or so maybe we can find something that is more natural online um so then when we think back about the functions of ambiguity I was talking about that it can increase perceptions that people are yeah still doubting their opinion that they are not not certain yet so which leaves room for more opinions to exist and secondly also that it allows people to to show that they are trying to take the other person into account that they don't want to offend the other person and then I was thinking that maybe another way to do this is simple as this maybe expressing your opinion as an opinion can also accomplish this as compared to expressing your opinion as effect because when we looked at the online discussions a lot of people were just um presenting their opinion as if it was effect like that is not true that's not the case while in the face-to-face conversations there were more of those statements of I think and in my opinion etc so I thought maybe this very minimal manipulation can also work and this is very uh preliminary preliminary data so difficult word um so uh keep silent about this it still needs to be replicated but I'm doing that now so in this study um which is really recent uh we had participants read fake news articles from New Pantanal that's a online news site a website thing where people can also discuss about the news article in the comment section below it um so we made fake news articles with fake discussion sections underneath them and we had participants evaluate these discussions and we had two conditions so in one condition the participants read the discussion section where all the discussing people for all the discussions were using words like I think and in my opinion etc that was one condition I think in the opinion as opinion condition and the opinion as fact condition all these words were gone so there was no uh in my opinion or uh according to my thoughts or so so that was all just gone um and then we asked them whether they thought these people were having uh agreement where they were being in agreement and whether they were having a good relationship and also whether they would themselves feel yeah prone to also participate in this discussion um and I have two uh examples of the stimuli we use for both conditions so maybe you can tell me which one is which the first one that's the opinion as opinion so uh there we included those statements in in my opinion and I've heard enough uh uh I always think so it's very uh sort of personal so it's just my experience and my opinion and here it's just gone so the comment is also a little bit shorter that's a limitation of course but still the difference there's no anymore uh no other differences so this is the only difference um so and then to the results uh so we saw uh that this manipulation was effective so um we see that when people read a comment section where everybody expresses their opinion as if as if it is effect so without statements like I think they perceive less agreement between the interaction partners and also less harmony in this discussion um and that's interesting because as you saw the only difference between these two conditions was these little sentences of I think and in my opinion and also as a consequence of of this perceiving this agreement um and harmony people felt also more likely to join when they read the discussion that was uh where everybody expressed their opinion as an opinion so it also seemed to create a more of a yeah a pleasant space to also participate in the discussion um yeah so that brings me to the conclusion also of the second part of the lecture so we first looked at yeah why can we why why why uh does online escalate more easily than face-to-face um and so we saw that that was due to uh lacking responsiveness and lacking ambiguity so these diplomatic behaviors which can lead to misperceptions of uh disinhibition and disagreement and then we looked in the second part at how we might improve online discussions based on this knowledge so we saw that by inserting responsiveness from a face-to-face conversation in a text chat that might help to make it seem more uh more harmonious and nice um but inserting uh ambiguity of the face-to-face type uh was not very effective that was a bit weird but then expressing one's opinion as an opinion seems to be a viable alternative for this um and because you're here also maybe to to maybe get some uh uh things tips to to go home right to also maybe improve your own uh online life uh I also included some uh take home tips for how to maybe behave in uh online discussions next time you engage uh so how can we improve uh online discussions and first of all I think it's very important uh that you think before you you act of course it's always important but especially here because we saw that there can be a lot of misunderstandings between people so maybe uh the the person you meet online and that seems so aggressive and anti-social maybe this person uh is not uh intending to behave in this way but is just limited in their behavior by the medium um and then secondly yeah always try to be responsive when you react to a online message so try to acknowledge the previous uh post that you say like ah yes but or so that you make a connection to to a previous statement um and lastly uh yeah maybe express your opinion as an opinion rather than as a fixed fact might also help so that was uh my presentation and these are my their collaborators my promoters so thank you are there any questions yes then please raise your hand if you have a question you have a question I was wondering to what extent the format or the platform where the discussion takes place has influence on um the tone of voice and the expressions that you just told about and do you do you uh sorry can I ask some more do you have some expectations about this what's can happen what might matter for example on reddit uh texts seem to be more um longer or so yeah more elaborate and on discord where there is a certain community people can have short texts or maybe on facebook short comments or preferred over long comments did you see anything on that or um yeah we didn't study this uh but I can imagine that when you're limited in the text you can use it becomes even more clear so that might lead to even more escalation so I know on twitter you have to be very short so maybe that can contribute I'm not sure but it is funny because we also we did look at reddit comments but then just at reddit comments and comparing it to others um but there we saw especially that people didn't really react to each other so there was um maybe somebody started to talk about COVID and then also that um their their mother was in the hospital or so and then somebody else reacted to that by asking by by talking about the hospital or so like oh yeah this hospital is shit or something like that so they just seem to be taking uh elements from each other's messages to react to so to take the other person's comment as an excuse to say what they wanted to say anyway so this can also maybe make people feel very ignored and yeah not really hurt so yeah that's interesting as well but thank you for your question all right and the others I see one in the back just a second hi um so I have I actually have two questions but my first would be um so in your research did you um kind of like explore different um places on the internet and how they differ per levels of what I would say toxicity because I've noticed that the way that for instance to bring it on myself that I interact on twitter is probably a lot more different than I would interact on say reddit or another social media platform and so I then begin to think that maybe some places kind of harbor a particular kind of behavior where you're like okay the area they're like the you're on twitter so therefore you're more combative for instance yeah and so like I was just wondering if you've looked into different um places yeah it's a good question um yeah we also didn't study this uh but this um it is known in literature that uh certain platforms can have certain norms social norms and this so how to how people tend to behave in these places so what we did look at in reddit is different uh subreddits that have also different norms so there are subreddits where where people are uh know they can be quite aggressive or so that's like the accepted way to to act to be very um um yeah aggressive in your language use and others where people are trying to be very constructive in their discussions there's different norms that also steer behavior indeed and also uh these norms might also dictate the consequences of this behavior because when you expect to be in a place where everybody is calling names and being aggressive etc maybe this is also less hurtful because you think like oh yeah that's the expected behavior and this is how we do here right so this is actually a very important factor to also take into account yeah and um just my second question was um so during the actual experiment um were there like what were your limitations i'd say like what did you feel like maybe you could have done better or to give you know more like uh more insightful like results or something because like for instance from my perspective i was like i found it really interesting that you got a um like for instance an actor to be kind of like an instigator right um but then i would wonder like what if uh you kind of like interviewed people prior and knew where they stood on like a little bit on a political spectrum and then like put those two people together for a more organic kind of like thing but that's just my opinion yeah yeah it's a good point as well yeah yeah and we indeed introduced this confederate because we noticed in the first discussion or in the first study that people were not very strongly opinionated and not really disagreeing about the topics but i in the first study i also tried to select discussion topics that were controversial among students so we did pilot test them etc and we also tried to indeed put people together with different opinions but still people were a bit like yeah we are just here for for the assignment and the study points um so then we had this confederate to get a little bit more uh yeah i don't know yeah controversiality um but then of course a very big limitation of all lab studies is that it is a lab study so it's quite it's not how real life might be or it's like a condensed sort of way to look at it so um it's still the ecological validity is limited so therefore we need to look indeed also in these real online discussions in what happens there but that's can be quite hard as well because they're there then you don't know about how people feel so i would like to know how somebody might feel when they have these online discussions on Reddit or so but the the thing is that it's very hard to find those people and interview them then you can just observe their behavior with no consequences or no psychological consequences so yeah that's all has um advantages and disadvantages so yeah thank you for the questions are there any other questions maybe yeah i have another question um i i first i would like to be clarified which were the age groups in in these two studies that you conducted um which which age groups are i think the first study there were like students or something and in the second study also yeah also students because i have a question do you think that your results are generalizable for the general population and if not which are your hypothesis or i better say expectations of implementing your studies to different age groups yeah yeah it's also a very good question and also a limitation of the studies that we looked only at this younger age group but i think that these are more general processes at least for the diplomatic behaviors we see that yeah in a literature that people often yeah are responsive and ambiguous as well in offline conversations so we know that and i also think that the consequences of of this behavior lacking might be still hurtful also in older generations um but it might be a less or uh or more pronounced for certain age groups and then i'm also thinking about uh yeah something that might be a little bit related to norms or so so that you have maybe um yeah related to expectations so older people might be more used to um yeah to offline uh uh uh conversations so maybe they might be yeah less understanding that the online medium can change behavior and feel even more hurt or not uh i'm not sure yeah so i find it hard to um to think about how it might be different can go both ways for my feeling but i'm not sure whether you have another hypothesis i personally think that younger like students like us are more prone for a conversation because um i might get into another field of nostalgia and stuff but things things are very different nowadays we we have already encountered conversation in social media about difficult topics yeah and um yeah so that's that's why i think in like older age groups are more hesitant this is but this is my own um uh hypothesis so yeah i think that yeah but when you think about uh i'm not sure what you mean with hesitant also the ambiguity side of it because when i'm now thinking about when i send was it messages to my mother she is quite clear and sometimes i feel a bit like oh is she angry or so a little bit so she also is less um yeah i don't know practiced in maybe adapting a little bit to the medium apparently so at least i also try to maybe include some emojis or just try to include some happy words or so not just clear and then with a with a dot at the end so maybe also having this limited experience might also make them appear even more disinhibited or angry or so so yeah that could also happen i think yeah good question there's one more i just i thought of something related to what he said um i think older people tend to be more settled in their opinion especially when you look at like the age group that is now about 18 so gen z they are more likely to listen to input from other people at least in my experience that might be my corner of their internet chiming in but i in my experience adults are older adults are very settled in their opinions and their thoughts and they don't listen to changes and then accept them um which i think might also make them if they are adapt in their abilities in online discussions make them more disinhibited or make them appear more disinhibited because they are so settled and they do that thing of like stating opinions as facts because to them it is a fact it is not an opinion yeah might be uh yeah going home interesting thank you all right any others yeah um so i had a little bit of a question i thought of when you were showing the take home points uh you mentioned limitations of medium yeah and i was wondering so both of your studies kind of focus on variables that are more language focused like ambiguity in terms of oh i think or stuff like that and responsiveness which is also more like the actual words um i was wondering what you thought about tone because often when you usually online conversations are text based and it's very difficult to deduce the tone someone's talking and especially because we all have different textiles and if you're trying to interpret someone's message just based on how you type it sometimes it's vague that way so i want i'm just wondering what you think because i can imagine misinterpreting someone's tone can easily escalate a conversation because you just think they mean something different yeah yeah i think yeah that certainly can happen indeed um so uh yeah this was also something i was struggling with when i wanted to do these studies because we did something that's a little bit maybe weird in a way in that we compared to very different media so we compared offline with online this is like a world of difference and and many things that could be coded in an offline conversation such as facial expressions or tone of voice or all these non-verbal things that are by definition not present online so i couldn't yeah when i would code it it was zero or so so then you don't have any variants you could work with so yeah in that way i i decided for now to leave that out but certainly it will be very important when you indeed smile with your message or say it in a happy tone or so or maybe a little bit like a questionable tone or so like i'm not sure something like that so that that will definitely have a huge effect yeah very good point yeah all right thank you for that um any others that still want to add something i don't think so all right then i would like to thank you karla um yeah very interesting to see this overview and also yeah see the research that you have done i think i learned a lot so that's that's good thank you and also thank you for all your very engaging questions that's always nice when you yeah when you think of that so thank you for the the audience as well please again a very big applause for karla rose