Originally posted on The Horizons Tracker.
Expertise has had a bit of a rocky time of late, with both various political figures deriding their talents and indeed various studies highlighting their seemingly scattergun success rates. Couple this with the rise of automated decision support tools and one might imagine experts to be getting a little nervous.
In an approach that mirrors the ‘augmented intelligence’ nomenclature preferred by IBM, the researchers analyzed the estimates of experts in agribusiness at Dow AgroSciences. The human experts did indeed manage to improve profits by 3% and decrease costs by 6%.
“Every year, the company needs to figure out how many acres of land they are going to use to produce seed corn,” the authors say. “But in this competitive industry, many varieties of the seed corn are new, and the company does not have a lot of experience in growing the new type. As a result, it does not know what the yield would be, or how many bushels of corn they will get from its fields. Yet, an estimate of the yield is necessary to optimize the resources used for growing seed corn.”
Much of the criticism of experts is down to the biases and relatively limited mental models they use when deriving their conclusions. This creates a very unequal terrain in terms of expertise, with considerable challenges involved in determining who actually are the best experts, and indeed by how much they are better than their peers.
The researchers gathered predictions from a number of experts about the quantiles of the yield. A model was then developed to convert this estimate into mean and standard deviation of yield.
“The mean provides estimates for how many bushels the firm can expect on average, while the standard deviation captures the expected variability in the growth process,” the authors explain.
The predictions of each expert was then compared with historic data to understand the variance in the work of each expert, and the biases they brought to the task. The model allows for a more realistic comparison of the merits of an expert, whilst also providing context to the organization regarding the suitable times to hire experts, or even the best ways to train their own.
The team hope to apply the model to other industries in future, with an initial focus on the biofuel and semiconductor industries, with both chosen due to the high level of supply uncertainties they deal with.
Article source: New model assesses the merits of experts.
- Bansal, S., Gutierrez, G. J., & Keiser, J. R. (2017). Using Experts’ Noisy Quantile Judgments to Quantify Risks: Theory and Application to Agribusiness. Operations Research. ↩