🔗 doi.org/10.1016/j.leaqua.2025.101941
Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk
End 🧵
08.03.2026 12:57
👍 0
🔁 0
💬 0
📌 0
🔗 doi.org/10.1016/j.leaqua.2025.101941
Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk
End 🧵
💡 The study recommends: When selecting or evaluating leaders, look beyond impressive communication. A polished charismatic speech may seem impressive – but if it truly requires cognitive ability to craft, it becomes harder to fake and more meaningful as a leadership signal. 6/🧵
➡️ Many leadership behaviours like expressing gratitude, personal disclosure, or wearing certain attire are treated as credible signals in research – but they can often be easily copied by anyone, regardless of actual leadership quality. 🤔 5/🧵
➡️ The study argues that leadership research fails to properly address signal costs – meaning most studied "signals" may not actually help followers distinguish good leaders from bad ones. 😬 4/🧵
➡️ For a leadership signal to be credible, three conditions must be met: the quality it reveals must be unobservable, the signal itself must be observable, and the signal must be costlier to fake for those who lack the quality. 🎯 3/🧵
A review study by @schimmelpfennig.bsky.social, Charles Efferson, and Nicolas Bastardoz takes a close look: 2/🧵
#Wilmasreview
Wilma 🐶 knows a thing or two about signaling – a wagging tail means she's happy, and paws on a big pipe means business. 💪
But how do we know when a leader's signals are the real deal? 1/🧵
➡️ Emotions are not just a side effect of AI use — they are a central driver of how people interact with technology. Balanced AI regulation should keep this in mind. 🤳
🔗 doi.org/10.1080/10447318.2025.2594748
Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk
End 🧵
➡️ The number one strategy researchers have studied for sparking emotional responses? Making AI more human-like. Think chatbots with personality, voices with warmth, or robots that smile back. 🤖😊
3/🧵
The study shows:
➡️ Most research focuses on how our emotions shape whether we use AI and how satisfied we are with it — but far less attention goes to what triggers those emotions in the first place. 🔍
2/🧵
#Wilmasreview
Wilma 🐶 has a lot of emotions when it comes to the dishwasher. Is it a source of leftovers? Or will it take her job?
Humans may have similarly complicated feelings about AI systems — and a review by König and Mehrotra maps out a whole decade of research on that. 🤖💭
1/🧵
💡 Policymakers can't rely on majority opinion here. Those most directly affected — welfare claimants — need a seat at the design table.
🔗 doi.org/10.1038/s414...
#postdog 🐾 Wilma, PhD
@au.dk End/🧵
➡️ Distrust in AI welfare decisions spills over into distrust of government overall. And wrongly denying benefits drives that distrust far more than slow processing speed. 📉 3/🧵
➡️ People who depend on welfare benefits are significantly more skeptical of AI decision-making than those who don't — even when the AI is faster. 🧐
➡️ Non-claimants overestimate how willing claimants are to accept AI in welfare systems — and can't correct this even when incentivized. 🤔 2/🧵
#Wilmasreview ❄️ Wilma isn't looking for a new home in the snow — but she is wondering: should AI decide who gets welfare benefits like housing support? A new study by Dong, @jfbonnefon.bsky.social & @iyadrahwan.bsky.social looked at exactly this. 1/🧵
➡️ Simply informing people about benefits still helps, but the real barriers lie in completing paperwork and navigating bureaucratic processes. 📋
🔗 doi.org/10.1002/pam....
Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk
End 🧵
➡️ Applying for benefits is not equal to receiving them – it is easier to help applying than actually getting benefits. 📝
➡️ Providing hands-on assistance (reducing compliance demands) increases take-up strongly – more than double the effect of just sending information. 🤝
2/🧵
#Wilmasreview
Wilma 🐶 met a snowy friend today! ❄️ This poor bear tried to apply for social benefits to get some food in this cold weather, but could not receive support despite being eligible.
A meta-analysis by @karlemilbendtsen.bsky.social reviews 51 field experiments to find out what works:
1/🧵
It must be very hard to publish null results Publication practices in the social sciences act as a filter that favors statistically significant results over null findings. While the problem of selection on significance (SoS) is well-known in theory, it has been difficult to measure its scope empirically, and it has been challenging to determine how selection varies across contexts. In this article, we use large language models to extract granular and validated data on about 100,000 articles published in over 150 political science journals from 2010 to 2024. We show that fewer than 2% of articles that rely on statistical methods report null-only findings in their abstracts, while over 90% of papers highlight significant results. To put these findings in perspective, we develop and calibrate a simple model of publication bias. Across a range of plausible assumptions, we find that statistically significant results are estimated to be one to two orders of magnitude more likely to enter the published record than null results. Leveraging metadata extracted from individual articles, we show that the pattern of strong SoS holds across subfields, journals, methods, and time periods. However, a few factors such as pre-registration and randomized experiments correlate with greater acceptance of null results. We conclude by discussing implications for the field and the potential of our new dataset for investigating other questions about political science.
I have a new paper. We look at ~all stats articles in political science post-2010 & show that 94% have abstracts that claim to reject a null. Only 2% present only null results. This is hard to explain unless the research process has a filter that only lets rejections through.
🔗 doi.org/10.1093/qje/...
Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk
End/🧵
➡️ The study recommends skills-based assessments when selecting managers, rather than relying on who puts themselves forward. The best leaders might be the ones who aren't raising their paws! 🐾
4/🧵
➡️ Good management isn't predicted by personality traits, demographics, or desire to lead – but by economic decision-making skills such as the ability to allocate resources wisely. 🧠
3/🧵
➡️ People who eagerly self-promote into management roles may actually perform worse than randomly assigned managers – likely because they tend to overestimate their own abilities. 😬
2/🧵
#Wilmasreview
For Wilma 🐶, figuring out how to recognize a good leader was always like solving a riddle of the Sphinx. A recent study by Weidmann, Joseph Vecci, @farahsaid.bsky.social, Sonia Bhalotra, Achyuta Adhvaryu, Anant Nyshadham, Jorge Tamayo, and David Deming suggests...
1/🧵
➡️ Such effects can persist for at least two weeks after for example a presidential election, and the patterns were easier to reactivate once they'd been triggered. 📅
🔗 doi.org/10.1287/orsc.2024.18538
Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk
End 🧵
➡️ Elections can reduce social mindfulness toward politically different colleagues. After elections, employees showed less perspective-taking and empathic concern toward coworkers with opposing political views – often leading to more negative interactions. 😔
4/🧵
➡️ Political differences between coworkers don't always cause friction – but around elections, things can change. Before the U.S. elections, political dissimilarity had no significant effect on negative interactions. But on election day and days after, it did.
3/🧵
A study by Max Reinwald, Rouven Kanitz, Peter Bamberger, Prof. Dr. Julia Backmann and Prof. Dr. Martin Hoegl examines this – how political differences affect workplace interactions, and why timing matters.
2/🧵
#Wilmasreview
Wilma 🐶 had an unexpected encounter with a deer the other day – and instead of running away, they just stood there peacefully! 🦌
It got Wilma thinking: why can some "natural opposites" get along just fine, while others struggle?
1/🧵
🔗 doi.org/10.1111/psj....
Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk
End🧵