Repair the Loop

The sixth principle is the one that matters most: detecting a pattern changes nothing by itself. A system improves only when observation is connected back to a deliberate correction, and when that correction is tested rather than assumed. One change, applied, checked. One, not five.

What the science says

The discipline of testing corrections is codified in the Model for Improvement (Langley and colleagues, The Improvement Guide), adopted by the Institute for Healthcare Improvement. It pairs three questions (what are we trying to accomplish, how will we know a change is an improvement, what change can we make) with the Plan-Do-Study-Act (PDSA) cycle: plan a small test, do it, study the result, act on what you learned. A correction is a hypothesis to be tested on a small scale, not a permanent decision made on faith.

But not all feedback helps. In their landmark meta-analysis The Effects of Feedback Interventions on Performance (Psychological Bulletin, 1996), Kluger and DeNisi reviewed 607 effect sizes and found that more than a third of feedback interventions actually made performance worse. Feedback that draws attention to the self (blame, identity, ego) backfires; feedback that keeps attention on the task and the next concrete action is what drives improvement.

Inside Loguaron

This principle lives in the Monthly Strategy, where you choose one main problem, not five, and define the next correction. The single-problem rule keeps the loop small and testable, exactly as PDSA prescribes, and keeps attention on the task rather than personal failure, exactly as Kluger and DeNisi demand. If a correction worked, continue it; if it failed, replace it. The question it answers: What is the one correction this month?