What Is a Dynamic Hypothesis?
Dynamic Hypothesis: A Testable Explanation of Behaviour Over Time
A dynamic hypothesis is a proposed explanation of how a system's structure generates a pattern of behaviour over time. It identifies the feedback loops, stocks, flows, delays, decision rules, and constraints believed to be responsible for the problem under investigation.
The hypothesis is dynamic because it must account for change, not merely correlation at one moment. It should explain why growth accelerates, why performance oscillates, why a policy loses effectiveness, or why a condition persists despite repeated intervention.
Development usually begins with a reference mode describing the historical, expected, feared, or desired behaviour of important variables. Evidence from data, documents, theory, interviews, and direct observation then informs a causal account capable of producing that pattern.
Causal-loop and stock-and-flow diagrams communicate the proposed structure. Equations and parameter values make the explanation precise enough to simulate. If the resulting model cannot reproduce the important qualitative behaviour, the hypothesis or its implementation requires revision.
A dynamic hypothesis is not a claim that one structure explains everything. It is purpose-specific and bounded. Competing hypotheses should be considered, and the model should be tested against evidence not used to formulate it, extreme conditions, dimensional consistency, and sensitivity to uncertain assumptions.
The concept turns modelling into scientific inquiry. Rather than assembling variables until a curve resembles historical data, the modeller proposes a causal mechanism, derives its consequences, compares them with observation, and progressively improves the explanation.
Back to reading