Schappert, W. Holzbauer, J.P. Pischalnikov, Y.M. Sergatskov, D.A. JACoW Geneva, Switzerland 978-3-95450-178-6 10.18429/JACoW-SRF2015-TUPB091 English 814-818 TUPB091 cavity coupling resonance factory superconducting-cavity Contribution to a conference proceedings 2015 2015-12 http://dx.doi.org/10.18429/JACoW-SRF2015-TUPB091 http://srf2015.vrws.de/papers/tupb091.pdf Measurements of cavity quality factor measurements are subject to at least three potentially large sources of systematic error that have not been previously recognized. Imperfect coupler directivity (cross-talk) can lead to large errors in the cavity coupling factor when the cavity coupling factor is significantly different than unity. Energy re-reflected from the circulator can systematically bias the measured cavity decay time which is used to determine the loaded quality factor. Use of the peak probe power or the minimum of the reflected power to determine the cavity resonance frequency rather than the peak of the probe/forward transfer function may lead to errors in the resonance frequency that can also affect quality factors. Each effect is illustrated with measurements in the Fermilab VTS, simulations and analytic calculations. If the magnitude and phase of the cavity RF signals are measured, these effects can be measured and corrected for. If only signal magnitudes are recorded or these effects are not measured, they must be treated as sources of systematic uncertainty.