We studied the influence of the statistical properties of soil moisture changes on the Interferometric Synthetic Aperture Radar (InSAR) coherence and closure phase to determine whether the InSAR coherence and closure phase can be used to estimate soil moisture changes. We generated semi-synthetic multi-looked interferograms by pairing n real single-looked pixels of an observed SAR image with n synthetic single-looked pixels. The synthetic SAR data are generated from the real SAR data by applying soil moisture changes with a pre-defined mean and standard deviation of changes. Our results show that the diversity of soil moisture changes within the multi-look window gives rise to decorrelation, a multi-looked phase artifact, and a non-zero phase triplet. The decorrelation and closure phase increase by enlarging the diversity of soil moisture changes. We also showed that non-soil moisture changes can lead to larger decorrelations and closure phases. Furthermore, the diversity of phase changes, decorrelation, and closure phases are correlated with land cover type. We concluded that the closure phase and coherence are independent of the magnitude of soil moisture changes and are inappropriate tools to estimate soil moisture changes. Coherence, however, can be used as a proxy for soil moisture changes if the diversity and magnitude of soil moisture changes within a multi-looked pixel are strongly correlated.