Theory: How Do You Motivate Individual Performance on a Distributed Human-Computer Team?
PARC alum Alan Kay is quoted as saying “A point of view is worth 80 IQ points.” This means that how we look at a problem makes a difference. A point of view is a theory.
As innovators, how should we determine which theories to use?
This post is about the human part of human-computer teams and how different theories of motivation lead to different designs for digital nervous systems.
The previous blog post suggests that the way to improve team performance is to provide the information or recommendations that people need when they make decisions.
But what if they don’t want to do their jobs?
This question is not simply rhetorical. Enforcement officers are often out of sight of supervision for much of an eight-hour shift. Not all officers are diligent in their work activities. In some cities, TV news organizations on slow news days have filmed officers sleeping in their vehicles or going home for hours during their shifts. These stories have played on the evening news. This has created urgency in finding ways to improve organizational performance and reduce slacking.
Going Beyond the Numbers
In designing CitySight®, we worked with people over a wide range of reported performance. We found that some officers value the public safety and public relations mission of the parking enforcement organization more than its revenue or compliance-related mission. Consider the following scenario:
An enforcement officer checks a vehicle parked near a grocery store. The parking meter has just expired. He notices a woman pushing a shopping cart and rushing quickly toward him. A baby is riding in the shopping cart with the groceries.
What should the officer do? (a) Quickly issue the citation and tell the woman to watch her parking times more carefully. (b) Help the woman load her groceries in her vehicle. Then send her off with a reminder to try to watch her meter time. An officer concerned with “letter of the law” and enforcement performance might choose option (a). An officer concerned with “spirit of the law” might choose option (b).
What Motivates Performance?
The upcoming releases of CitySight will have more real-time data collection and analysis. But how should the analytics and alerts be designed to best improve performance? There are several competing approaches about how to do this. The different approaches rest on different hypotheses and theories about what is most effective and sustainable.
Here are three broad approaches.
- Coaching in the moment. This approach uses monitoring software to detect when officers are not being productive. The theory is that low performance is caused by slacking. The hypothesis is that performance will improve if supervisors get real-time alerts about low-performing officers and coach them when it is happening.
- Real-time recommendations. This approach uses predictive analytics to create recommendations for optimal performance. The theory is that low performance is caused by people not knowing what to do. The hypothesis is that by providing recommendations just in time, officers, dispatchers, and supervisors will be informed, make better choices and improve performance.
- Motivational interfaces. This approach is about making jobs more interesting. Researcher Daniel Pink summarizes a body of psychological research on motivation and performance. (The cartoon for this post is inspired by a video explaining his work.) For any activity that involves even a little bit of cognition, the greatest performance is found when the task provides a sense of autonomy, mastery, and purpose.
Innovating from Theory
Our innovation team recognized that the three approaches are complementary. They can be combined in several ways. As we thought about this, several more dimensions and questions became apparent. If a system provides real-time recommendations, should it provide them to officers directly or to their supervisors? If officers were organized in teams, would peer pressure have a positive or negative impact? We created a matrix of dimensions of design and discovered many combinations. We noticed that different combinations had different failure modes. For example, if people are not interested in their jobs, they might resent or ignore recommendations. At the same time, there might be ways to provide feedback and shared cognitive puzzles that would increase the experience of mastery.
If we just picked and deployed only one approach, it would probably work for some organizations and some situations, but not for others. Or it could work for some part of an organization, but not for others. Poor results would be frustrating for our clients.
So we are designing the system to enable organizations to learn what works best. It should be configurable so that organizations can try different strategies and collect data about performance. This is the course we prefer for upcoming releases of our system.
Evolving Practices and Theories
In closing, by trafficking in both theory and practice innovators contribute to both. Like our clients, we are interested in understanding how to create very high performing human-computer teams.
It has not gone unnoticed that our own journey of innovation is exciting for us in part because it provides us with a sense of mastery, autonomy, and purpose.
My group at PARC is called TAO, which stands for Theory and Technology for Agile Organizations. We design and implement web services that help organizations optimize their activities, using mobile and cloud technology. Thanks to the CitySight team and our partners for all of the discussions and design that informed this project. Thank you to Matt Darst, Hoda Eldardiry, Bob Krivacic, Lawrence Lee, Raj Minhas, Sai Nelaturi, Mudita Singhal, and Barbara Stefik for comments on earlier versions of this post.
Our work is centered around a series of Focus Areas that we believe are the future of science and technology.
We’re continually developing new technologies, many of which are available for Commercialization.