Profound Knowledge Pt 3: Theory of Knowledge

How do we know what we think we know is really so?

In my earlier article on Understanding Variation, I described how data can mislead if we don’t understand the patterns and causes behind it. Let’s revisit that idea with a common scenario:

A manager reviews a performance chart and notices a dip. Without understanding variation, they jump into action — changing processes, shuffling staff, or tightening controls. But the “problem” was actually just natural fluctuation. Instead of improvement, the changes add instability and frustration.

The difference between this reaction and a measured, effective response is knowledge — not just information.

From Information to Knowledge

Deming emphasized that true knowledge requires a theory to explain what we observe and predict what will happen next. He outlined three building blocks:

1. Experience – The lessons, observations, and outcomes we’ve encountered before. It’s where most of us start when interpreting results.

2. Theory – A structured explanation of why something happens. A good theory can be tested and challenged. Without theory, experience stands alone and can easily mislead.

3. Prediction – The ability to forecast results based on theory. If predictions consistently fail, the theory must be refined or discarded.

Information is not knowledge.

A list of data points (e.g., “last month’s defect rate was 5%”) is just information. Understanding why it was 5% and being able to forecast next month’s rate — that’s knowledge.

The Role of Psychology

Deming also recognized that human beliefs, motivations, and biases shape how we interpret information. One of the most common is confirmation bias — the tendency to seek and trust only information that supports our existing beliefs. Left unchecked, this bias leads to flawed conclusions and poor decisions.

PDSA: Putting Knowledge into Practice

One of the most practical tools for applying the Theory of Knowledge is the Plan–Do–Study–Act (PDSA) cycle:

1. Plan – Form a theory and design a change or test.

2. Do – Carry out the plan on a small scale.

3. Study – Compare results to predictions.

4. Act – Adjust the theory or standardize the change.

By following PDSA, we link theory with evidence and prevent changes based on hunches or one-off events.

Avoiding Tampering

Without a sound theory and understanding of variation, well-intentioned managers often “tamper” — reacting to isolated data points in ways that actually increase variation. The result? Wasted effort, lower morale, and poorer outcomes.

The Insight

Problem solving is complex. Don’t rely on experience alone, and don’t let cognitive bias lead you to disrupt a stable process.Instead:

• Build a theory.

• Test it.

• Compare results to predictions.

• Learn and refine.

This is how organizations move from reacting to reality — to shaping it.

For more on Deming’s System of Profound Knowledge, visit the Deming Institute