In an experiment with 206 law students, AI highlighting of legal complaints cut task time by 29.7% with accuracy unchanged, while AI summaries saved no time and combining both tools was slower than highlighting alone.

Using pitch-tracking data and police reports for nine MLB home counties from 2009 to 2019, an economist found domestic violence reports rose after losses in games with highly inaccurate umpires, driven by missed calls that favoured the opponent.

Smartphone tracking of up to 1,229 University of Texas students linked political identity to 18 of 61 everyday behaviors, mostly in leisure, but fellow students misjudged 76% of those behaviors and overestimated the divide.

In an Oxford University Press chapter, psychologist Jens Gaab argues that psychotherapy and open-label placebos are close relatives, noting that therapy’s specific advantage over placebo controls in depression is small and only marginally above the threshold for a clinically meaningful difference.

Between 2012 and 2022 the ideal number of children in rich countries fell from 2.42 to 2.31, and plans to have a child within three years from 28% to 21%. The fading belief that children are needed for fulfillment contributed most.

An archaeologist identified 659 diagnostic and probable dice at 57 North American sites, including 20 from 12,000-year-old Folsom deposits, predating the earliest known Old World dice by more than 6,000 years.

An economic model shows why influencers endorse only on-brand products while building awareness, then switch to promoting almost anything once a large share of potential followers know them. Celebrities sell out sooner, and viral platforms can flip the pattern twice.

Managers who regularly answer unfavorable analysts on earnings calls see those analysts soften their views, get stronger market reactions to their forecasts, and suffer smaller stock losses when a peer firm restates its accounts.