Due to the exponential attenuation of photons in materials, thick samples will attenuate a large portion of photons. This is a source of error in methods such as neutron activation analysis, which use gamma spectroscopy to characterise a radioactive source. A method is developed to quantify the magnitude of self-shielding with the help of MCNP6. Then the mass of an unknown sample is determined by comparing the known sample mass and the photopeak counts of the unknown and the known samples following activation. The inclusion of self-shielding effects is shown to make this comparative mass analysis measurement technique more accurate. Accounting for the self-shielding effects allows the true source, instead of the shielded source, to be resolved by correcting for the photons that are attenuated as they try to escape the sample and reach the detector. The $\gamma$-ray measurements were made using several samples of varying shapes.