Biologically inspired engineering design requires understanding of complex biological systems for use as analogues in engineering designs. In this study we seek to understand how functional representations, in particular Structure-Behavior-Function (SBF) models, enable understanding complex biological systems. Results from this study indicate that SBF representations may enable more accurate inferences about biological systems for complex and abstract questions than purely textual, or textual and diagrammatic, representations. They also suggest that no one representation is best for all types of inferences.