166 - Steve Rathje: The Psychology of Virality

27:33In addition, if you look at the research on gossip, which goes back to the 1940s in social psychology, there's a strong literature on gossip. It seems like similar types of gossip go viral as well. Gossip is disproportionately negative. It's almost always negative. And it's often moral in nature. We're often discussing the moral transgressions of others. So it seems like it's not just moral outrage that goes viral on social media, but moral outrage goes viral in terms of gossip. And I know we use this term viral often to discuss social media content, but this paper we introduced the term viral to talk about the spread of information broadly.

28:12Information has always gone viral in some senses in our offline networks after the invention of the printing press. The term viral just seems to refer to the tendency for content to spread widely. One thing we do in this article is we review some of the key studies on what goes viral, both on social media and in our offline networks in terms of gossip. And we plot all these key studies on according to their arousal and their valence. So a lot of emotion theorists talk about emotion in terms of valence or how negative an emotion is, and how high arousal or intense emotion is. So we plot all these key studies on these two axes, negativity and arousal.

28:58Most studies support the idea that what goes viral is often both negative and high arousal. Oftentimes they are a combination of highly negative and high arousal. So for instance, outrage, outgroup animosity, anger, these are all high arousal and negative emotions. However, high arousal negativity isn't the only thing that goes viral and it seems like that there are a lot of context dependencies some studies suggest that high arousal positive content goes viral for instance analyses of the most shared articles in the new york times suggest that emotions like awe or high arousal positive emotions go viral one analysis of data from upworthy suggested that people were likely to click on articles that evoked sadness so a low arousal negative emotion So there are a lot of inconsistencies, but there are some clear patterns.

29:51Interestingly, we didn't find any articles suggesting that low arousal positivity ever predicted virality. So emotions like calmness or calm never seem to really go viral. So there are some patterns on what goes viral, and there seem to be some context dependencies. And some of these patterns seem to be rooted in key psychological processes. So for instance, we know from studies on attention and memory that what grabs our attention and what we tend to remember seems to be both negative and high arousal. So one reason that negativity and high arousal content might go viral is because it simply captures our attention.

30:31In a crowded information environment, high arousal negativity is what we're going to attend to and what we are going to remember. So it seems like these psychological processes might explain why content goes viral on social media as well as in our offline networks via gossip. However, we also aim to reconcile some of these conflicting findings in the literature. Why does positivity sometimes go viral? Why does high arousal positivity go viral sometimes? Why does low arousal negativity go viral? Why do these conflicting findings exist? One reason, of course, is that psychological data is often messy and social media data is messy.

31:08But another reason is likely because of structural factors of information environments. So in the second half of the review article, we talk about how structural features of an environment, such as algorithms, incentive structures, or norms or networks also shape virality. You talked about attention and motivation. These are mostly related to the self. We are motivated or our attention is taken. But sharing information, it's not only about grabbing attention or engagement, but it's also a deeply social act shaped by motivations like identity, status, reputation, affiliation, and the desire for a shared reality.

31:54I'm really interested in how these social motifs can sometimes interact with or override other considerations like accuracy. How do you think this tension plays out when individuals decide whether to share, modify, or withhold information? And more broadly, how should we think about the difference between me as an individual consuming information privately versus me as a social being embedded in a network responding to and acting on that same information? Yeah, great question. What I talked about earlier were primarily cognitive processes that drive information sharing, what we attend to, what we tend to remember.

32:38And I think that's certainly part of the story. Something that grabs our attention is more likely to go viral. On TikTok, for instance, they pay close attention to watch time statistics. And what you tend to watch, what you can't look away from, is what goes viral. But as you mentioned, there are a lot of social motivations at play as well. When we are choosing to share an article or a piece of information or a piece of gossip, this is a self-conscious social motivation. We're thinking about our self-presentation. We're thinking about our social identity. There are neuroscience studies, for instance, that suggest that when you view viral content, it activates areas in the brain that are implicated with social processing.

33:19It seems like when we're thinking about what to share and when we're looking at viral content, this is often a really social decision. We call it social media for a reason. There are a ton of articles that look at the various social motivations that drive sharing behavior. There are articles on motivations for status driving social sharing behavior, motivations for belonging, motivations for conformity. There's a great article suggesting that one reason we share misinformation is because this conformity desire. We'll have this fear of missing out if we don't share misinformation that promotes our group.

33:53We do it for this social reason. There are also sometimes antisocial reasons why we share articles. There's great work by Michael Bang Peterson that suggests that one of the key predictors of the sharing of misinformation and hostile political rumors is a need for chaos, basically. However, I call this an antisocial motivation, but it's also a status-seeking motivation because Michael Bang Peterson finds it's really people who have this belief in what's called a need for chaos or this desire to burn the system down, who share misinformation and hostile political rumors. And often these people have this desire because they are frustrated with the system and they think that this is a means of them for gaining status.

34:35If they are constantly trying to generate chaos, this is perhaps a means for them to get status. Indeed, he finds that individuals who are highest in status-seeking motivations have this strong need for chaos and often share misinformation and hostile political rumors. You also mentioned how accuracy and social motivations, there can be a tension between them. And I think that's what my study that I mentioned earlier that looked at whether you experimentally manipulate accuracy and social motivations, how that shapes sharing behavior. And indeed, we found in that study that if you make social identity motivations particularly salient, it can often distract from accuracy motivations and cause you to share or believe in misinformation.

35:19And we think that this is oftentimes what's happening on social media because social media makes the social aspect of sharing hyper salient. You get a lot of rewards. You get likes and shares. Social media isn't necessarily an environment that makes accuracy salient. There are other environments where accuracy is more salient. If you're a scientist, for instance, you're embedded in a system where you have peer review. You have people constantly critiquing your work, and also your identity is sort of intertwined with accuracy. Part of your identity and your reputation depends on being accurate.

35:56That's not necessarily true on social media, and especially in some political communities on social media, where perhaps it's more important to signal your partisan identity than necessarily to be accurate all the time. we tackled the psychological processes that drive information, dissemination, or sharing behavior but one of the key ideas you explore is the information as a virus metaphor answering the question one you also touched upon that in this review you mentioned both its strengths and limits I would love for you to dig into it a bit more for us out of the studies you review could stand without this metaphor, yet it plays a central role in how the paper frames information spread.

36:46So what do you think makes this metaphor especially compelling to use? And where do you think it breaks down, especially when we consider the role of human agency and how intentional so much information sharing actually is? yeah one of my first interests in psychology was actually in the psychology of metaphor going back to that intro psychology class that i took my freshman year one of the most interesting studies i learned about in that class was work by lara borodisky that found that if you frame crime using the metaphorical frame of a virus it changes how people think about crime versus if you frame crime as a beast.

37:31And you can instantiate this metaphorical frame using just a few words. But if you propose that crime is a virus that is infecting this hypothetical town, people tend to propose more systemic or reformative solutions to deal with crime. But if you say that crime is a beast that is haunting this hypothetical town, people tend to propose more punitive solutions. George Lakoff wrote this book called Metaphors we live by back in the 1980s that was basically about how we oftentimes use metaphor to think, and we often need metaphor to think and to reason through things. I think that the virus metaphor is really unavoidable if you look in this literature.

38:14It has really been so central. We talked about viral videos in the early 2000s, but people have been using this metaphor of information as a virus for decades. Metaphors can help and harm. Sometimes they really help shed light on an issue, and sometimes they can change how we think about an issue, as I talked about with that prior study. And I will say there are some notable critics of this information as a virus metaphor. And we cited a number of the notable critics of this information as a virus metaphor. Some people think that maybe it's too negative when, you know, information is something that's pretty neutral.

38:49Perhaps it implies that people are too passive, for instance, and that people are passively infected by viruses when psychology is really more complicated than that. People can ignore information, people can change information. So this metaphor has flaws, but I also think it has a number of strengths, and I think it's pretty powerful. And I also think that all metaphors have flaws as a thing. You're never going to find a perfect metaphor. But yeah, beyond metaphor, people have even used epidemiological models to model the spread of information. And these models aren't perfect. No model is perfect, but to some sense gives us some way to think about information spread.

39:29In the paper, we review some ways in which this information as virus metaphor is useful. So like viruses, some forms of information are more contagious than others. People talk about viruses having a certain are not value, which is a measure of how many people the virus is on average thought to infect. So just like how some viruses have a higher R0 value or are more contagious than others, it seems like some forms of information are more contagious than others and have always been more contagious than others, like negative information or moral information. In addition, information exposure has psychological consequences.

40:06It changes us. Information, like viruses, can mutate over time, and some mutations of that information might become more contagious over time. During COVID, we talked a lot about COVID super spreaders. There were super spreader events. There's also research suggesting that some people on average are more infectious than others when it comes to viruses. We don't exactly know why, but there are some people who tend to be super spreaders of viruses. And there also tend to be super spreaders of information and misinformation. There's a really powerful statistic that suggests that around 0.1 of Twitter users share about 80 % of misinformation on Twitter.

40:51So there's this huge sort of Zipion distribution where it's only a small handful of people who are sharing a ton of misinformation. And you also see similar patterns when it comes to toxicity on Reddit. It's only around 3 % of active users who contribute to about 33 % of toxicity. So you see these distributions where there are these super spreaders of false and hostile content on social media. In addition, just like how some viruses will spread faster in some contexts, so like a crowded room where people are close together and where they're en masse, a virus is going to be more spreadable, some types of information are more spreadable in some contexts.

41:35And this gets to where I was talking about the structural features that drive the spread of information. There will be some networks or there will be some social media platforms where misinformation or hostile content are more spreadable. In addition, many other people have talked about interventions that inoculate people against misinformation, essentially. Just as there are ways to stem the spread of illness and viruses, there are ways to stem the spread of viral misinformation or polarizing content. I think this metaphor is quite powerful and can be extended in length, and we use it as a central frame for the paper, while also acknowledging its downsides.

42:17Metaphors can be great for theorizing, and they can help us theorize, and they can help us think, but we should also not let them lead us astray. So it's a powerful metaphor that we should also take with a grain of salt. one of the most intriguing findings you identify is the paradox of virality the idea that widely shared content is often not widely liked and i think this is a central theme in some of your previous work as well this seems counterintuitive for content that becomes so prevalent um what are the primary explanations you propose for this paradox because as i mentioned it just doesn't on your intuitive that people oftentimes make the content that they tend not to like go viral.

43:04What do you think it is? Yeah, what led us to study the paradox of virality is Facebook when describing their algorithm, they will say publicly on their website, this is how we predict the content that people want to see. And I was thinking after publishing this work on outgroup animosity, I was like, do people really want to see outgroup animosity? I don't think they do. People might engage with outgroup animosity, but I don't think they like it. And sort of an interpretation of my paper at that time was people were like, oh, that's human nature. People love to hate on the outgroup. And I was like, do they really love to hate on the outgroup?

43:40I mean, people do it. People engage in outgroup animosity, but I don't think that's what people like. So we conducted a study where we surveyed a nationally representative sample of United States participants, and we asked them what do you think goes viral on social media and what do you think should go viral on social media in an ideal world? And we found that people were actually pretty aware that divisive and negative content tended to go viral on social media. People's perceptions of what went viral were in line with past research on what actually goes viral. However, we found that people expressed a strong stated preference such that they thought that divisive and negative content should not go viral, and that instead educational or scientific or positive or nuanced content should go viral and be amplified by social media algorithms.

44:27So then that leads to the question, if people don't like this content, why do they engage with it? Why do they literally click like and share when it comes to divisive content? And there are three potential explanations we propose for this paradox. One is just that people's revealed preferences are different from their stated preferences. This language of revealed and stated preferences comes from economics, and a lot of economists will say that we should kind of ignore people's stated preferences and just pay attention to their revealed preferences. So that's one interpretation of this finding.

44:58You can just say that people's stated preferences of what they want to see on social media are unimportant. People engage without group animosity, therefore they like it. But I don't think that that's a correct interpretation of this finding, because I think when it comes to things like addiction, for instance, we should pay attention to people's stated preferences. For instance, there's research suggesting that around 70 % of cigarette smokers say that they want to quit smoking. But we don't say that they have a revealed preference for smoking. We'd instead say that they're addicted to smoking and that they don't want to continue.

45:30And I think it might be similar with social media. It might be like, we don't actually like this content, but instead we can't look away. In some way, we're addicted to it. It's a little bit like if you see a car crash on the side of the road and you stop to look at it. it's not because you like that car crash or want to see car crashes in the world it's just that you can't look away yeah that's one potential explanation basically is that it's not that people like this content but it's that it's attention grabbing and because social media is an attention economy that tries to be addicting and have us use their product again and again they won't show us content that we actually like or say we that we want to see they'll show us content that we can't look away from.

46:12And I think that's one explanation for the paradox of virality. But I think another explanation that is often ignored or not paid enough attention to is this idea of super spreaders that I talked about earlier. I don't think it's the vast majority of people who are liking and sharing outgroup animosity on social media. I actually think it's a small percentage of people. People have done survey experiments to follow up on some of my work on outgroup animosity and they find in these experiments that if you survey just like a representative sample of u.s participants people won't show this preference for sharing outgroup animosity in a survey experiment or they won't show a preference for sharing divisive content however when you zoom in on the small percentage of participants who are highly active on social media and who engage a lot with politics on social media they actually say that they do have a preference for outgroup animosity, and they do want to share outgroup animosity.

47:13So I think that's what's going on as well. When you log on to social media, it's a few hyperactive super spreaders who tend to be more extreme. They tend to like more polarizing content, and they're sharing a lot of the divisive content on social media. There's really an invisibility of moderate voices on social media, and this can have negative consequences. It can have us infer false norms or hold false beliefs about the opposing party because sort of the only representation that you'll see of the opposing party or even of your own party are the most extreme voices because those are the people who are most motivated to post on social media.

47:51I remember about those super spreaders, one of the co-founders of this podcast, Eric Newman, he had a recent study about that. And I remember him going around in the lab, just 3%. They're just like representing 3%. That's amazing. And it stuck in my mind. That's so interesting to hear. And oftentimes people are unaware of these facts. If they knew, I know maybe our behavior towards these posts or these information, maybe it would be different. Yeah, yeah, yeah. I know some of his work in this area. And I think he's trying to educate people about these super spreaders. And I think he finds that that's generally effective.

48:29So I think that That is one route to fixing some of the issues we have with social media is either I talked about my unfollow study earlier, you can unfollow those super spreaders, you can clear up your feed and your information diet, or you can just be more aware and have some literacy about these super spreaders. Your paper also outlines how structural futures of information environments such as algorithms, platform design, these platforms norms and network size and properties interact with these psychological processes. And in your review, you also mentioned historically much of the research has framed the comparison as online versus offline information environments.

49:09But now with the expansion of both offline and online spaces, each with distinct structural characteristics, it seems the landscape is far more complex now. How do you see these varying structural differences shaping the ways people engage with information across different contexts? And in what ways might this challenge or expand our traditional understandings of information dissemination and our behavior towards these information? So yeah, your point about how the online and offline world is increasingly interconnected is, that's a good one. And that's one we grapple with a bit. It can be so hard to study social media because, you know, more than half the world is on social media and it has such complex effects, basically.

49:57It doesn't just affect individuals. It affects communities. And it's really hard to even isolate online and offline behavior because what happens online informs our offline discussions and vice versa. It's super interconnected. In this paper, we try to zoom in on a few structural features of social media environments and online environments and explore what their impact might be and how their impact might be different than offline environments. I emphasize the online and offline similarities. People are people wherever they go. hostile people tend to be hostile online and offline. We tend to be attracted to negativity in all contexts.

50:34So there are important similarities, but there are also important differences. One difference is algorithms, for instance. You don't have algorithms mediating your in-person conversations, basically. I say that, but then I immediately think that there are things like dating apps that use algorithms to filter who you talk to in real life. So that's one important nuance. But basically, when you're having an in-person conversation with someone, your messages aren't being filtered by algorithms and i i think that's a key difference algorithmic audits of twitter's algorithm for instance have found that what tends to be shown in the algorithmic as opposed to the non-algorithmic feed on twitter tends to be more negative about the outgroup tends to be more polarizing as expected especially in a political domain and tends to not align with people's stated preferences.

51:23Interestingly, this doesn't seem to be the case for non-political tweets. The algorithm seems pretty good at showing us effective non-political tweets, but when it comes to politics, it will usually show us the most divisive posts. So that's one important structural feature that's really, really hard to study, actually, because we don't really know how algorithms work under the hood. A lot of the engineers of the algorithms also don't know how they work under the hood. But there have been great audit studies, like the one that I just mentioned, some of the ones that we study, that try to get at it.

51:54Another structural feature is design features of social media platforms. So one thing that's interesting about Twitter, as opposed to Reddit, for instance, is that you have a like button and that you have a retweet button, but you don't have a dislike button. And that probably informs what goes viral, because the algorithm can't have negative signals of what people dislike or what people don't want to see. Jeremy Freimer has some great work where he finds that the most liked posts on Twitter tend to be paradoxically the least liked. And he also finds that they tend to have the lowest like per retweet ratio.

52:36So basically, what happens here is really divisive or toxic content gets retweeted a lot. And as a result, it gets shown to more people and it gets more likes, but it has a lower like to retweet ratio indicating it's less liked. And he also had independent reviewers review these posts with the low like to retweet ratio. And he confirmed that the like to retweet ratio seems to be a good indicator of whether or not someone actually likes or approves of a tweet. So this phenomenon happens on Twitter, basically, where these tweets that people don't like will just be highly liked because they're highly seen.

53:17They're so retweeted, they drive a big conversation because they're very polarizing. So adding a dislike button could potentially change those dynamics, for instance. So that's an aspect of a design feature. Other researchers have found that if you added a trust button or a misleading button to a social media posts, or if you actually had someone rate the accuracy of a post and you added that as an algorithmic input, that would absolutely change social media. So that's another key design feature that's important. Another important structural feature is the norms of various social media platforms.

53:52The norms of LinkedIn are very different from the norms of TikTok. And we have preliminary research suggesting that negativity doesn't seem to go as viral on LinkedIn. It seems to be more about arousal as opposed to negativity. Some people talk about toxic positivity on LinkedIn. There isn't really a norm about being negative or engaging in outgroup animosity, whereas I think the norms of Twitter, especially after Elon Musk purchased Twitter, became quite divisive and polarizing. So that also shows how the norms of a social media platform can rapidly switch. The norms of Twitter rapidly switch after Elon Musk purchased Twitter.

54:26Another important structural feature is network size, or network size and structure, I would say, because those are both important. One key difference between social media and the offline world is there are hostile individuals online and offline, but online hostile individuals have huge networks and they can spread misinformation and toxicity around the world rapidly in just a few minutes amplified by attention-grabbing social media algorithms. Now, there are also super spreaders in the offline world, I was looking at the literature on gossip, and there are super spreaders of gossip as well. There are people who gossip way more than others.

55:05But what's interesting about the gossip super spreader literature is it suggests that while a little bit of gossip predicts increased friendship, it will predict you having more friends, people who gossip the most actually have the least amount of friends. It's a negative predictor of friendship. And this is the opposite of what happens on social media, where spreading rumors and toxicity and hostility will get you a lot of followers. If you're more extreme, you have a lot of followers. But if you're like that in the offline world, that won't get you a lot of attention. So those are four structural features that we talk about algorithms, design features, norms, and networks.

55:42It's so interesting to see that we see all these patterns looking at different platforms, different cognitive processes, different people who are engaging with different mediums. I think what you're doing here with this review is pointing us to a direction saying that how we study it doesn't matter because we see all these things happening across platforms, across mediums, online or offline. Thank you. At the end of your review, you talk about some of the future direction. You mentioned that most of this literature is conducted on English language data in Western world. There are not that many cross-cultural references happening.

56:23Looking ahead, how did this review direct you? Or where did this review direct you? How do you describe the paper? I liked how you described the paper because we wanted to be really big and grand with this paper. We wanted it to be a big review that's all-encompassing of virality across history, across context, across online and offline, around the globe, for instance. I think one key important point for the future of the field is looking at virality on a global scale. As you mentioned, there are very few cross-cultural studies. This is a problem that's inherent to social science and a lot of psychology in general is that we don't have enough cross-cultural research and most of our samples are limited to a United States context.

57:10And there's some preliminary research that suggests that what goes viral in other contexts might be very different. For instance, there's work conducted by Jeannie Psy at Stanford. Instead of viral, they talked about information being contagious, which they operationalize as more likely to impact people and change how they communicate on social media. So they looked at something slightly different from virality, but it's related, I think. So basically, they found that high arousal negativity was contagious in the United States, But in more collectivist cultures like Japan, it was actually high arousal positivity that was more contagious.

57:48And one potential explanation for this stems from Jeanne's size theory of ideal affect, different cultures prioritize different emotions. So for instance, individualist cultures like the US prioritize high arousal positive emotions like excitement, whereas more collectivist cultures tend to prioritize low arousal positive emotions. So the fact that we value and express emotions differently in different cultures will probably mean that we'll get lots of results in different cultures. And I thought that work was very inspiring. And it inspired a project that I thought of with my advisor that we call the global virality study.

58:25We thought of it actually a few years ago, where we really wanted to look at what goes viral in countries around the world. And we really wanted to look at cross-cultural differences and country-level moderators, like individualism, collectivism, perhaps moderating what goes viral in different countries. And what's interesting is that when we initially thought of this idea, this was before large language models took off. And some of our methods for doing text analysis were a lot worse. It was a lot harder to analyze text from multiple countries around the globe. But in 2023, I published a paper on how GPT is a really, really effective tool for analyzing text, analyzing psychological constructs in texts in multiple languages.

59:12We found that GPT was highly effective at detecting sentiment, emotions, and other psychological constructs in 12 different languages, including a lot of lesser spoken languages. So now that we have large language models in AI, we have a lot of potential to analyze what goes viral around the globe and look at these cross-cultural differences. And that's what I'm trying to do with some of my current and future work. And also, I think AI unlocks some interesting new methods such as agent-based models that can simulate people and and simulate groups of people, and simulate societies, and what goes viral in groups of people.

59:50I mentored James Hay on this paper where he was using little societies of large language models interacting with each other to basically simulate collective behavior. And he found that not only do large language models seem to simulate individual behavior, but they also simulate collective behavior, like our tendency to form echo chambers or engage in homophily. So we He's now left academia. He's founding a startup called Artificial Societies that aims to do this on a really large scale and simulate giant societies to look at how messages spread within communities. I really think AI has just unlocked our methods and will lead to sort of a revolution in our methods of studying virality to the same scale, or perhaps even a bigger scale, to what the internet did to our ability to do computational social science.

1:00:40I'm really excited for AI-based methods for the future of virality. Thank you so much for giving us all these insights into your paper. Before we leave our conversation, I just want to ask one more question. For students or professionals who are interested in both psychology, social psychology, and technology, what advice would you give them for bridging the gap between these two fields? Are there key skills or knowledge areas they should focus on? And also like Stanford psychology podcast, it's basically a science communication platform. And I know that you're also big on science communication.

1:01:21And how do you think that is helping you do science? Yeah, great questions. Okay, so first, my advice for students who are interested in the psychology of technology, my advice might differ depending on whether you come from a psychology background or a more technical background like computer science. My advice for people who don't maybe come from a psychological background or are steeped in psychological theory is I think I have really benefited from basing all my work and all my questions pretty deeply in psychological theories. Like social identity theory, like motivated cognition, naive realism.

1:01:57I really base my questions about how we interact with technology on the psychological literature and psychological theory. And I think that really helps us create better questions and do better studies. It gets you much further than just saying, what is the impact of AI or what is the impact of social media to really zoom in on a psychological theory. And then my advice for students who are really already steeped in psychological theory is get their hands on new methods. What I really benefited from early in my PhD was really taking the time to get effective at computational social science and analyzing big data, learning these new methods, your PhD is probably really the only time in your career where you'll be able to sit down and learn all these new methods.

1:02:46And some of my broad advice too is pick your advisor carefully. Advisors really, really matter. In academia, your PhD and postdoc advisor, I'm very lucky with the advisors I've had who have been absolutely incredible, who let me be both independent, but also steered me in the right direction and gave me really amazing advice at all times. I'll connect this advice question to my work on science communication, because my background in theater and playwriting and writing was so transferable to science and really, really helped me both in science and science communication. And sort of toward the end of my time as an undergrad, I began writing op-eds, for instance, The Guardian or Psychology Today, and I was writing about the work of Rob Willer or the work of J.

1:03:33Vaughn Babel, who ended up being my postdoc advisor, and that really helped me think through ideas and communicate ideas well. And learning to communicate is so important, and science communication, it helped me form better ideas, and it helped me write scientific articles better. And how you communicate your idea is so important to having your idea go viral or spread widely. Like communication is so important. And academic writing can be so hard, it can be really dense. I think doing science communication really teaches you to be clear. And yeah, so also some more background, I've been running this science TikTok channel, Steve Psychology, for the past five years.

1:04:14I've been able to grow a following on that platform, and that also helps me with public speaking. It will help me with teaching in the future. I think doing science communication, it's often underappreciated in some departments, especially more traditional psychology departments. But I think it's really important growing your skill set in writing and communicating, which is so central to the field. And more than that, it's also just really valuable to see just how people outside of academia and outside of the ivory tower will benefit from their work. It's really nice to see people excited about my work and about psychology work in general and see how psychology ideas benefit people's lives.

1:04:58And I think a lot of us are in research, not just because we want to communicate to other academics, but we want our research to have an impact on the world. And I think one of the greatest ways our work can have an impact is through science communication and through spreading it to the world. Such a wonderful way to wrap up the episode. Thank you so much for joining me, Steve. Thank you. thank you so much for listening if you enjoyed this podcast help us make even more people excited about science by leaving us a review on spotify apple podcast or elsewhere and subscribing to our no spam all fun substack at stanford scipod to connect with other listeners you can also shoot us an email with your thoughts or suggestions at stanford scipodcast at gmail.com thank you and have a wonderful day

1:05:47Thank you.