Where Data-Driven Decision-Making Can Go Wrong

Harvard Business Review (Sep-Oct 2024 issue, vol 102 no. 5) by Harvard Business School professors Michael Luca and Amy C. Edmondson argues that business leaders default to one of two errors with data evidence — treating findings as gospel or dismissing them outright. The remedy is rigorous discussion of internal validity (does the analysis answer the question?) and external validity (do the results generalise?). The authors flag specific traps: confounding causation with correlation, ignoring sample size, setting and period, and measuring what is easy rather than what matters. They couple analytical rigour with the soft skill of psychologically safe deliberation — directly relevant to SEA mid-market board meetings where data slides go unchallenged.