Featured Product
This Week in Quality Digest Live
Management Features
Dawn Bailey
The focus is on preparation, communication, and inclusion
Julie Winkle Giulioni
Old givens are giving way to new ungivens
Gartner
Few hybrid workers report feeling connected to their organization’s culture
Huw Thomas
A long-awaited expansion of workers’ rights
Gleb Tsipursky
Quality professionals should test their intuitions before deploying internal and external quality initiatives

More Features

Management News
Sept. 28–29, 2022, at the MassMutual Center in Springfield, MA
EPM service provider excels in helping customers work with EPM products
It’s not exactly a labor shortage
Program inspires leaders to consider systems perspective for continuous improvement and innovation
Recent research finds organizations unprepared to manage more complex workforce
Attendees will learn how three top manufacturing companies use quality data to predict and prevent problems, improve efficiency, and reduce costs
More than 40% of directors surveyed cite the ability of companies to execute as one of the biggest threats to improving ESG performance
MIT Sloan study shows that target-independent compensation systems can be superior

More News

Amir Goldberg

Management

How Machine Learning Can Help Leaders Make People-Related Decisions

Machines make the predictions; but it’s you who makes the decisions

Published: Thursday, June 16, 2022 - 12:02

The first step in using data is understanding what data analytics can and can’t do. Artificial intelligence (AI) systems are powerful but are best used for prediction models. The role of a leader is to use those predictions to inform decisions.

I’m a professor of organizational behavior at the Stanford Graduate School of Business, where I teach a class titled People Analytics. People analytics is the application of data analytics methods, especially machine learning algorithms, for the purpose of informing people-related decisions inside organizations. Here are five takeaways from that class.

1. Be an informed consumer of data analytics

Data are revolutionizing how people are managing inside organizations. Using data analytics doesn’t mean that you need to be a data analyst. What you need to be is someone who understands what data analytics can and can’t do. Most important, understand how to interpret data analytics in terms of what it means about what’s happening inside your organization.

2. AI-driven doesn’t mean AI drives

It’s really important to understand what algorithms can or can’t do. People call it artificial intelligence, but machines aren’t intelligent in the way that humans are. What are they good at? Prediction, by which I mean looking at historical data and then making predictions about the future. Your role is to use those predictions as a means to inform your decisions. It’s the machines who make the predictions; it’s you who makes the decisions.

3. Be skeptical

There’s a lot of snake oil out there. If it sounds fantastical, it’s probably no good. It’s very difficult to predict human behavior. The most important thing for you to ask is whether the prediction is about an outcome that you actually care about. If the person who developed the algorithm can demonstrate to you that it moves the needle on outcomes that are important for you, then it’s worth considering. But if they can’t, you should probably pass.

4. Use algorithms to de-bias your decisions

When it comes to making decisions about other humans, all of us are biased. We pay attention to other people’s race, to their gender, to their physical appearance, to their accent. Those are often entirely irrelevant to the decisions we need to make. So use your algorithms in a way that would help you overcome those biases and make your decisions better and more ethical.

5. Be ethical

Decisions about other people can be immensely consequential. They can relate to their livelihood, to their sense of worth, to their psychological well-being. We can’t take those lightly. It’s really important that when you make these decisions, you don’t hide behind machines, and you don’t outsource the morality of the decision to a machine-learning decision-maker. Ultimately, it’s your responsibility. Machines are neither moral nor immoral. They do what you tell them to do. It’s important that you think about the moral implications of your decisions. Fortunately, ethical decisions are also better managerial decisions. Good luck.

First published May 10, 2022, in Stanford Graduate School of Business Insights.

Discuss

About The Author

Amir Goldberg’s picture

Amir Goldberg

Amir Goldberg is an associate professor of organizational behavior at Stanford University whose research lies at the intersection of cultural sociology, data science, and organization studies. As co-director of the computational culture lab, he uses and develops computationally intensive network- and language-based methods to study how new cultural categories take form as people and organizations interact.