Your digital footprint may reveal more about your personality than you realise. A 2024 meta-analysis of both human perception and machine learning studies finds meaningful but imperfect accuracy, with humans performing at least as well as algorithms on average.
What actually changes what people do? A comprehensive synthesis of meta-analyses finds that the most popular intervention strategies, targeting knowledge and attitudes, are the least effective. The approaches that work best are those that make desired behaviours easier to enact.
Does hybrid working damage performance? A six-month randomised controlled trial with 1,600 employees finds it does not, while reducing quit rates by a third. The managers in the study revised their own views after seeing the evidence.
Personality feedback is one of the most common tools in leadership development and coaching. A systematic review finds only 12 studies examining whether it actually improves performance, and the evidence for causal benefit remains unestablished.
Algorithms consistently outperform human judgment when combining assessment data to predict performance. A 2023 survey finds that practitioners know this and largely ignore it, and identifies the specific reasons why the evidence has failed to change practice.
Job stress has long been linked to heart disease. An 18-year study of 6,000 workers finds that when job strain and effort-reward imbalance combine, the risk more than doubles, with implications for how organisations think about workload and recognition.
Research on coaching effectiveness has identified what matters. A 2024 systematic review goes further, specifying what effective coaches actually do and, crucially, when they do it across the phases of a coaching engagement.
Cognitive ability tests are a staple of graduate recruitment. A 2023 critical review finds that the validity evidence used to justify them has been overstated, their adverse impact is real and consequential, and their use in high-volume screening raises questions that the field has been too slow to confront.
How reliable are supervisor performance ratings, the standard measure against which all selection tools are validated? A 2024 meta-analysis finds higher reliability than previously estimated overall, but meaningful variation by job type that current practices do not adequately account for.
When a psychology study finds that people judge X worse than Y, how many people actually did? A 2023 paper demonstrates that the answer is often far fewer than the finding implies, and that the gap between group statistics and individual reality is one of psychology's most underacknowledged problems.
