The influence of error structure on interpretation of impedance measurements is demonstrated for electro-hydrodynamic impedance spectra. The stochastic component of the error structure was determined through use of measurement models, and a three-parameter model for the error structure was identified. The measurement model was also used to identify portions of the impedance spectra that were not corrupted by bias errors. Regression employing weighting by the stochastic component of the error structure yielded unambiguous values for physical properties such as the Schmidt number, whereas significant ambiguity was observed using modulus or proportional weighting.