A new piece in JLSP evaluated how sexual orientation and age categories are conflated. Check out the paper here! journals.sagepub.com/doi/full/10.11…

How do adolescents linguistically portray migrants? Check out new research in JLSP! journals.sagepub.com/doi/abs/10.117…

I'm excited to share a new paper published in JPSP APA Journals! In short: when people value effort, complex language can be a positive heuristic that links to behavior in the field. This idea is revealed in over 1M online petitions across 8 languages. psycnet.apa.org/record/2023-26…

This work builds on the research that Hillary Shulman and I published in PNASNews on goal instrumentality and processing fluency. Check out these papers! pnas.org/doi/abs/10.107…

In a new preprint, jeff hancock, Stanford VR, and I evaluated the differences between AI-generated text and human-generated text with experiences (writing about a hotel stay, news). AI text is more affective, analytical, and less readable than human text: psyarxiv.com/mnyz8/

What does verbal analytic thinking really measure? New work finds that the most reliable link to verbal analytic thinking is need for cognition, and less so cognitive ability. Content effects also matter! See here, in Applied Cognitive Psychology. onlinelibrary.wiley.com/doi/10.1002/ac…

New paper on disciplinary bias in Educational Researcher - Editors. Using NLP on 3.5M comments, teachers sending students to the office were more negative toward Black vs. White students, and more impersonal toward Asian vs. White students. journals.sagepub.com/eprint/6SNGIMV…

Just out in The Journal of Language and Social Psychology. Pushing back agains the pushback in news interviews. When intersubjectivity is invasive and coercive. Do we need to talk about interknowledgeability? doi.org/10.1177/026192… @MarissCaldwell

📣 Thrilled to share our (@LizStokoe & mine) latest The Journal of Language and Social Psychology #EMCA article on #resistance in #B2B ‘cold’ #sales calls 🥶☎️ 🗯️🙄😡 It’s🔓 #OpenAccess🔓& available here: doi.org/10.1177/026192… LSE Psychological & Behavioural Science VU Faculty of Humanities EMCA:tweets

AI cannot have experiences, but they can write like they did. This inherently false communication differs from intentionally false human communication in several linguistic ways (e.g., emotion). New piece with jeff hancock and Stanford VR in JLSP: journals.sagepub.com/doi/10.1177/02…

Now OPEN ACCESS! 🎆 >Pushing back agains the pushback in news interviews. >How intersubjectivity can be invasive and coercive. >And... interknowledgeability? doi.org/10.1177/026192… The Journal of Language and Social Psychology @MarissCaldwell

Are generative AI good deception detectors? We find that across LLMs, prompts, and texts, AI are much more truth-biased than humans. This may modify how we think about the origins of the truth-bias and the utility of AI for detection. With jeff hancock: psyarxiv.com/hm54g/

How do accents affect performance and potential biases during deception detection? We examine observers’ perceptions of four linguistic dimensions in our new article in The Journal of Language and Social Psychology journals.sagepub.com/doi/10.1177/02… Amy Leach Elizabeth Elliott

✨Excited to share a new paper with Vera Hoorens and Susanne Bruckmüller ✨ Does it matter whether we compare groups implicitly ("Women are brave") or explicitly ("Women are braver than men")? Yes, it does! A 🧵 The Journal of Language and Social Psychology journals.sagepub.com/doi/abs/10.117…

Generative AI are more truth-biased and truth-default than humans. In this new, four-study paper with jeff hancock, we present some of the first evidence comparing deception detection abilities of AI and humans. Check it out! The Journal of Language and Social Psychology doi.org/10.1177/026192…

Trending in #Linguistics: ooir.org/index.php?fiel… 1) The Bilingual Delay & Developmental Language Disorder 2) The barriers to reading for pleasure 3) Speech sound disorder or DLD (phonology)? (IJLCD) 4) Generative AI Are More Truth-Biased Than Humans (The Journal of Language and Social Psychology) 5)

NLP on spoken words predicts group engagement longitudinally in VR. Self & Collective references correlate with presence ratings. Impersonal speech correlates with large interpersonal distance. 130,000 words studied by Cyan & colleagues @JLangSocPsych. vhil.stanford.edu/publications/p…

Simplicity wins! 📰 In Science Advances, Hillary Shulman Todd Rogers and I show how online readers prefer simple news headlines over complex ones, w/ data from The Washington Post, Upworthy, and the lab. To capture attention, keep headlines simple for readers! science.org/doi/10.1126/sc…

New research in PNAS Nexus reveals that AI like GPT4 can simplify scientific writing, improving public understanding of science and perceptions of scientists. When AI crafts lay summaries, readers grasp content better and view scientists as more credible. doi.org/10.1093/pnasne…