Single and Double Loop Learning

in Improvement Science, Knowledge, PDSA

By David M. Williams, PhD

Simulation is a powerful method for teaching new improvers how to learn and use the plan, do, study, act (PDSA) cycle to test changes. In addition to the Mr. Potato Head and the Coin Spin PDSA exercises, there are a host of other worthwhile options like the sequence exercise. You can try a version here

In the sequence exercise, a set of numbers are presented – for example, 2, 4, 6 – and you are invited to try to figure out what rule the sequence obeys. To learn the rule, participants make a prediction about what it may be and then test another sequence that also obeys the rule. The facilitator confirms whether the tested sequence is correct or incorrect. Simple enough, right?

Without fail, most human beings will overthink the rule and come up with very clever but incorrect equations and alternative theories to explain the sequence. For example, the first number added to the second number equals the third number. To test this theory, one might try 4, 8, 12. The tested sequence does obey the rule, but the prediction of what the rule is (4+8=12) is incorrect. Additional tests may follow similar forms, even numbers (4, 8, 10), numbers going up by 4 (10, 14, 18), etc. Test after test often obeys the rule, but the prediction of the rule itself is wrong. 

The late Dr. Chris Argyris, professor emeritus at Harvard Business School, wrote a classic paper in the Harvard Business Review introducing the idea of single versus double loop learning. Single loop learning is when we note something doesn’t achieve the result we want and we alter or tweak our practice in hopes of getting a different result. This is often an adaptation of our existing practice. In the sequence exercise, the iterative tests of theories and number combinations are adaptations of the sequence shown as an example. The testing of similar sequences of numbers and confirmation they obey the rule limits the learning.

Now let’s imagine one of the teams suggests testing four numbers or a sequence of letters. While not a radical test, these sequences pivot from some previous assumptions that the rule requires three numbers. It challenges the underlying theory and may result in new learning. Argyris called this double loop learning and considered this a more comprehensive inquiry. Testing sequences that more strongly deviate from the initial theory provides deeper learning into what the rule may or may not be.

Figure – Single vs Double Loop Learning. Adapted from Maccoby et al. (2013). Transforming Health Care Leadership. A systems guide to improve patient care, decrease costs, and improve population health. San Francisco: Jossey-Bass.

How does this help you day-to-day as you’re on your improvement journey? When you note your practice is not producing the results you want, stop and ask yourself if you should make a change and adapt your practice in hopes of getting a different result (single loop). Then take a step further and ask, “What’s the theory behind our current actions?” Based on what you are learning and seeing, should you adapt your theory (double loop) in order to get a different result? How can you apply this more comprehensive inquiry to your work?

You can read Chris Argyris’s classic article at Harvard Business Review here.

If this was helpful, share and include me @DaveWilliamsATX. Sign up here to receive a monthly email from me that includes all my blog posts and other Improvement Science resources I think you’d appreciate.