Accidents and failures leave clues that appear to be clear in their meaning, but can be misleading if used without an informed investigation. A leak may be obvious, but the cause or moisture source is not. The cause could be due to a design deficiency, poor installation, materials or manufacturing defect, improper owner operation, or some combination of these. To sort this out and understand the cause with confidence, an expert with broad knowledge of codes and standards, design requirements, building component functions and limitations, installation methods, the performance and wear of materials, maintenance requirements, extensive investigation skills and experience, and project management skills.

A good investigation will systematically cut through the uncertainty of an accident or failure and bring clarity and authority. This is done by using the scientific method, which is the historic method for conducting technical inquiries and also is the standard set by the Supreme Court for expert investigations. The scientific method starts with making observations, not with conjecture about what happened. The observations include gathering information through field inspections, research on the design and installation, materials and their properties, maintenance requirements, actual maintenance and repair history, and the events of the accident or failure. Field and/or laboratory testing with specialized equipment or modeling may be required to supplement field observations. The observed data is then used to develop hypotheses of what went wrong. Hypotheses are tested based on facts in evidence to establish opinions of failure and responsible parties.

This sort of investigation may seem expensive, but - regardless of the complexity of the incident - this sort of investigation is actually far less expensive and far more effective than conducting an investigation based on unfounded hypotheses of what went wrong. Clarity and authority come from describing the foundational information that led to the hypothesis and by showing that the hypothesis predicts and explains what was observed in the accident or failure based on facts in evidence.