Adamson's System Analysis Tool (ASAT): A Systems-Oriented Framework for Recursive Experimental Planning, Dynamic Inference, and Scientific Interpretation

Scientific investigation has traditionally relied on controlled experiments that isolate variables to estimate causal effects. While indispensable, this approach becomes increasingly difficult in complex adaptive systems where multiple variables change simultaneously, causal structures evolve over time, latent influences remain unobserved, and experiments are conducted as part of ongoing operational environments. This paper introduces Adamson's System Analysis Tool (ASAT), a systems-oriented methodological framework for organising experimental portfolios, recursive scientific inquiry, and dynamic interpretation of evidence. Rather than replacing established statistical methods, experimental design, causal inference, or systems modelling, ASAT provides a higher-level structure that integrates these approaches within a recursive evidence-driven workflow. The framework introduces concepts including experimental portfolios, typed potential-relationship networks, recursive evidence ledgers, dynamic state estimation, latent-variable investigation, organised residual analysis, surprise and drift detection, rate-sensitive perturbation analysis, and adaptive experiment selection. Particular emphasis is placed on distinguishing observation, inference, simulation, and causal claims while supporting iterative refinement of scientific understanding. ASAT is presented as a conceptual methodology intended for empirical evaluation rather than a replacement for the scientific method. The paper explicitly identifies limitations, falsification criteria, validation pathways, governance considerations, and relationships to existing fields including systems thinking, system dynamics, design of experiments, causal inference, active learning, digital twins, Bayesian optimisation, and scientific experimentation. Potential application domains include agriculture, engineering, education, healthcare, organisational science, environmental systems, and other complex adaptive systems where recursive experimentation and systems-level interpretation are required.