Structural vs Statistical Thinking

Understand system behavior vs variability in pharmacometrics.
Tip

What you’ll build today: the ability to distinguish structure from variability.

Learning Objectives

  • Define structural thinking
  • Define statistical thinking

Key Ideas

Structure = system behavior
Statistics = variability and uncertainty

Both are required to make reliable decisions.

Worked Example: Same Dose, Different Outcomes

Two patients receive the same dose:

  • Patient A → high exposure
  • Patient B → low exposure

Structural view:

  • Same underlying model

Statistical view:

  • Differences due to variability

Insight

A perfect structural model can still fail if variability is ignored.

Conversely, focusing only on variability without structure leads to meaningless summaries.

Strategies

  • Separate signal (structure) from noise (variability)
  • Interpret both simultaneously

Common Mistakes

  • Ignoring variability when interpreting results
  • Treating variability as random error rather than meaningful information
  • Misinterpreting noise as signal
  • Overfitting structure to explain random variation
  • Ignoring uncertainty when drawing conclusions

Practice Problems

  1. Define both concepts

Structure = system behavior
Statistics = variability

Summary

PMx requires understanding both structure and variability.

  • Always separate signal from noise.
  • Variability contains information.
  • Both perspectives are required.