Multiscale entropy (MSE) is an algorithm to quantify the complexity of physiological dynamics system. It has been widely applied in different fields to probe into inherent information of biological signals. Although it is important for us to understand how these differences of signals patterns affect the trend of MSE, the factors impacting MSE and their mechanisms remain unclear. The aim of this study is to investigate difference of MSE for datasets with different sampling frequencies, amplitudes and waveforms. Our results demonstrate that:1) Increasing the sampling frequency will improve the resolution of MSE. 2) overall scaling of amplitude changes will not influence the MSE, while randomly amplifying the signal amplitude will influence the MSE significantly. 3) The waveforms would affect MSE hardly, especially at large scales. Our research of these affecting factors of MSE can help investigators to have a deepen comprehension of the algorithm of MSE and the meaning of multiscale when analyzing real world datasets.