Contrast-enhanced ultrasound (CEUS) imaging relies on distinguishing between microbubble and tissue echoes. Amplitude modulation (AM), a nonlinear pulsing scheme, has been developed to take advantage of the amplitude-dependent nonlinearity of microbubble echoes. However, with AM, tissue nonlinear propagation can also degrade the image contrast. Segmentation of CEUS images based on amplitude-dependent phase difference in the echoes, defined in this article as [Formula: see text], has been proposed as an additional method of enhancing contrast-to-tissue ratio as tissue is not expected to create the same degree of [Formula: see text]; however, this has not been robustly investigated. In this work, we evaluate the source of [Formula: see text] through simulations of unshelled versus shelled microbubble oscillation and simulations of nonlinear propagation in tissue. We then validate the simulated [Formula: see text] results with experimental [Formula: see text] measurements during in vitro scattering and imaging in a flow phantom. We show that shelled and unshelled microbubbles resulted in a [Formula: see text] with similar overall magnitude with some differences in trends, and that tissue echoes have a small yet detectable degree of [Formula: see text] due to nonlinear propagation. The results from this work can help inform optimal parameter selection for phase segmentation and implementation on a clinical scanner.