Via “Paul Krugman”: a paper, finding that most pundits do no meliorate at making predictions than a flipped coin. Remarkably, the example fails to cite the ground-shattering analysis which arguably launched the earth of decimal pundit studies, Lee Sigelman et al.’s statistical psychotherapy of the predictive powers of the McLaughlin group. Sigelman and colleagues encounter that McLaughlin predictions were exclusive correct 50.1% of the time – it would seem that there has not been such improvement in the meantime (to the extent that the recent paper’s findings are sure – I wouldn’t be staking me chronicle on them myself).
No comments:
Post a Comment