Range, David Epstein

• Breadth of training predicts breadth of transfer. The more breadth within learned content, the more the learner creates abstract models, and the less they rely on any particular example. Learners become better at applying their knowledge to situation they had never seen before—which is in essence creativity
• A study found that typical children tend to be raised in families with an average of six household rules, compared to one rule for extremely creative children. After a wrongdoing parents would let the child know, rather than proscribing it beforehand
• The “hypercorrection” effect: the more confident a learn is of their wrong answer, the better the information sticks when they subsequently learn the right answer. Tolerating big mistakes can create the best learning opportunities
  • This effect was even reproduced in primates (rhesus macaques)
• Teachers who drive overachievement in their current courses (by providing easier content) tend to undermine student performance in the long run (US Air Force study). Those who had more challenging math courses, and thus had lower student satisfaction, saw their students overachieve in future math courses (i.e., short term their grades would suffer, but long term their grades flourished)
• Interleaving problems, i.e., mixing them up when practicing, has been found to be more effective than approaching problems in blocks. Shuffle the problems beforehand to improve learning via rules differentiation rather than learning structured patterns
• According to a Yale study, more scientifically literate adults are surprisingly more likely to become dogmatic about politically polarizing topics in science. One theory is that they are better equipped to find evidence that confirms their beliefs
• A personality feature that fights back against this propensity is scientific curiosity, not scientific knowledge. Roam freely, listen carefully, and consume indiscriminately. Be open-minded.
• Scientific work that builds bridges between disparate pieces of knowledge (i.e., atypical citations with respect to the discipline) are less likely to be funded, more likely to be ignored upon publication, but more likely in the long run to gain traction and succeed