A model on the table gives a forecast, and an experienced manager opposite says the opposite. This scene is usually debated with the question of who is right; yet the right question is under what conditions algorithmic advice is taken into account. This episode builds the answer through three verified experimental studies. Dietvorst, Simmons and Massey (2015) showed that although evidence-based algorithms predict more accurately than human forecasters, people frequently choose the human forecaster, and they called this phenomenon algorithm aversion; the reason is that after seeing the same mistake, people lose confidence in algorithmic forecasters more quickly than in human forecasters, and this held even among those who saw the algorithm outperform the human. Logg, Minson and Moore (2019), on the other hand, found the opposite in six experiments: lay people adhered more to advice when they thought it came from an algorithm than when they thought it came from a person; the authors call this algorithm appreciation. The same study reports two boundary conditions: this tendency weakens when people choose between the algorithm’s estimate and their own, and when they have expertise in forecasting; the most striking finding is that experienced professionals who make forecasts regularly relied less on algorithmic advice than lay people did, which hurt their accuracy. Dietvorst, Simmons and Massey (2018) showed what resolves aversion: when participants could modify the algorithm’s forecasts, they were considerably more likely to choose to use an imperfect algorithm and performed better; this preference persisted even when the ability to modify was severely restricted, because what people sought was not more control but some control. It must be emphasized that this finding cannot be read as human intervention making the algorithm more accurate; the gain came from people starting to use the algorithm instead of their own weaker forecasts. It should also be noted that these studies were conducted on incentivized forecasting tasks, not on senior executives or real organizational decisions. In the final part of the episode, these three findings are read in an adaptation to aviation: knowing the system’s scope limits in advance, positioning the recommendation as an input rather than an order, and leaving the human a defined but meaningful area of intervention. At the end of the episode, you will clearly see that the leader’s job shifts from making decisions to designing how decisions are made.
📌 Playlist for the full series: https://www.youtube.com/playlist?list=PL6s4Uhec9cFBZNpSWXeuaybOuyCsHS71K
Sources Used
Dietvorst, B. J., Simmons, J. P., & Massey, C. (2015). Algorithm aversion: People erroneously avoid algorithms after seeing them err. Journal of Experimental Psychology: General, 144(1), 114-126.
Logg, J. M., Minson, J. A., & Moore, D. A. (2019). Algorithm appreciation: People prefer algorithmic to human judgment. Organizational Behavior and Human Decision Processes, 151, 90-103.
Dietvorst, B. J., Simmons, J. P., & Massey, C. (2018). Overcoming algorithm aversion: People will use imperfect algorithms if they can (even slightly) modify them. Management Science, 64(3), 1155-1170.
📚 Academic Use Note: This content is a supporting learning resource prepared with reference to published academic studies; it does not replace textbooks, lecture notes or your instructor's teaching. The sources to be used for exams and assessments are your course's official curriculum and the sources recommended by your instructor. The simplifications and interpretations in the narration are for teaching purposes; TT Institute Academy cannot be held responsible for any consequences arising from possible errors, omissions or outdated information.
© TT Institute Academy — All rights reserved.
