A new study from MIT Sloan School of Management identifies the most effective ways businesses can use and benefit from AI personalization. Researchers found that specific pairings of human and AI personalities can significantly improve ad quality and performance, with text quality increasing click-through rates and decreasing costs per...
MIT Sloan researchers demonstrate a new approach to designing social science experiments by applying integrative experiment design to a longstanding question about punishment in public goods games. The study reveals how punishment's effect on collective welfare can swing from substantially harmful to substantially helpful, depending on...
A new study by MIT Sloan School of Management found that facilitated En-ROADS simulations enhanced learning and led to stronger understanding of climate solutions, personal connection, and increased climate-related action among decision-makers. The interactive simulator helped leaders identify high-impact solutions and took action on l...
Researchers found that as systems grow, new functions are added at a slower rate, requiring infrastructure and existing functions to expand. Cities show an exception, where function diversity grows rapidly in smaller cities but slows down in larger ones.
A team of researchers proposes a novel adaptive platform trial fund to accelerate ALS drug development. The study outlines a financing model that merges efficiencies with an innovative royalty-based investment structure, drawing on data from the HEALEY ALS Platform Trial. This new approach could transform how diseases like ALS are addr...
A new study from the MIT Center for Collective Intelligence found that combining humans and AI can be effective in certain tasks, such as creative ones like summarizing social media posts or generating new content. However, human-AI combinations often underperform compared to AI alone in decision-making tasks. The researchers suggest t...
A new MIT Sloan research found that conservative users share misinformation at a greater volume than liberal users, explaining why conservatives are suspended more frequently. Conservative Twitter users shared four times more links to low-quality news outlets than liberal users.
Large language models can effectively reduce individuals' beliefs in conspiracy theories, with the effect lasting for at least two months. AI conversations challenged a wide spectrum of conspiracy theories, including those related to COVID-19 and politics, and led to shifts in behavioral intentions.
New research shows that social media fact-checking warning labels are broadly effective at reducing belief in and spreading of misinformation, even among those who distrust fact-checkers. The study found that warning labels can significantly reduce false headlines' accuracy ratings and willingness to share, regardless of participants' ...
A new study by MIT Sloan researchers found that misleading content from mainstream sources, rather than outright misinformation, was the primary driver of vaccine hesitancy on Facebook. This type of content was seen much more widely and generated more skepticism about COVID-19 vaccines.
Female STEM doctoral students are less likely to become new inventors compared to their male counterparts during PhD training. Faculty advisors play a critical role in training new inventors, with top inventor faculty members increasing the probability of filing the first patent by 23%. The research highlights a 'leaky pipeline' of fem...
Researchers found that Americans are more receptive to counter-partisan messages than previously thought, and party loyalty does not significantly distort their processing of evidence. The study used a survey of over 5,000 partisans to test the effect of persuasive messaging on their attitudes.
A new study published in Collective Intelligence journal suggests that evolutionary forces may be fueling collective tendencies to discriminate. The research recommends fostering environments where desired behavior emerges naturally through evolutionary dynamics rather than regulating against undesired outcomes.
The study found that identity cues accounted for up to 61% of the variation in voting, causing viewers to evaluate content faster. Positive feedback enhanced reputations, while negative cues led to down-votes. Reputations influenced how viewers perceived and engaged with online content.
Researchers at MIT Sloan School of Management studied 93 countries and found more than half willing to tolerate higher death rates without NPIs. The US, however, may face high daily death rates exceeding 1,000 per day during winter months if past responsiveness continues.
A large-scale experimental study by Harvard, Stanford, and MIT researchers found that weaker social connections on LinkedIn have a greater beneficial effect on job mobility than stronger ties. Weaker ties increased the likelihood of job mobility the most, while strongest ties had the least impact.
A new MIT model simulates vaccine allocation strategies to predict the effect on cumulative deaths and infections in the US. The research finds that allocating more than 50% of available doses to non-vaccinated individuals can significantly increase lives saved and reduce infections.
A new data-driven simulation model projects the national opioid crisis will worsen before it gets better, with over half a million additional deaths predicted over the next decade. The SOURCE model identifies three key strategies that could save more than 100,000 lives by 2032: fentanyl harm reduction, naloxone distribution, and recove...