The term was coined in 1953, two decades after the original studies, from a methodologically loose, largely uncontrolled dataset. A 2011 re-analysis of the original illumination-experiment data by economists Steven Levitt and John List found only slight long-run evidence for the effect, far weaker than its popular retelling, and the psychologist Richard Nisbett has dismissed the whole phenomenon as a glorified anecdote, so how much of the classic story survives contact with the original data is unsettled.
What would resolve this A full re-analysis of the surviving relay-assembly and bank-wiring datasets, not just the illumination-experiment data Levitt and List examined, using modern statistical controls for the original experiments' many confounds (rest breaks, pay changes, seasonal effects).
OpenIndustrial-organizational psychology and applied econometricsWikipedia: Hawthorne Effect