Neutron self-shielding has been a factor of concern in the history of Instrumental Neutron Activation Analysis. When the sample
to be analyzed cannot be made small enough in size and/or sufficiently diluted, this undesired phenomenon must be accounted
for. Several analytical, semi-empirical and computational methods for estimating the thermal neutron self-shielding effects
have been extensively discussed in the literature and this work aims at the experimental validation of some of these methods
by neutron irradiation of cylindrical samples containing strong thermal neutron absorbers. The accuracy and the relative differences
in the results between these methods are discussed for cylindrical samples with up to 40% thermal self-shielding, showing
that a semi-empirical sigmoidal function can be more accurate in modeling this effect than other exact algorithms, where a
maximum 2% relative difference to the experimental values was obtained.