Heat-generating magnetic nanoparticles suspensions are being explored in research and clinical settings as hyperthermia treatment for cancer or as adjuvant in established cancer therapies. In these applications it is essential to use low nanoparticle dosage to prevent any potential side effects including those associated with their accumulation in liver or spleen. Hence, developing particles with superior heating properties continues to remain an active area of research. Specific loss power (SLP), also referred to as specific absorption rate (SAR), represents the power dissipation per unit mass of magnetic nanoparticles in alternating magnetic fields. Accurate measurement of SLP is the key for understanding the parameters that control the heat generation rate, which is required to optimize these systems. However, at presents there are no standards for performing SLP measurements and no accepted calibration materials, making it difficult to compare the performance of various systems reported in literature. Previous work from this group discussed the effect of sample volume and geometry on the SLP data accuracy. In this study, additional analysis and experiments are carried out to investigate the effect of the power dissipation rate, the magnetic properties and the method for temperature slope calculation on the accuracy of the reported power density. Results indicate that when the same heating time is used, the volume at which heat losses become negligible decreases with decreasing sample heating rate. Furthermore, it is shown that for calculating initial temperature slope, a larger error occurs with a longer heating time and higher power level regardless of the curve fitting methods, hence, when power density or heating time increases, a higher order curve fitting (e.g., 2nd polynomial and exponential) is more desirable. In addition, when the magnetization of a nanoparticle suspension is low, the SLP is independent of the sample geometry.