Plasma polarimetry is a diagnostic technique used in nuclear fusion reactors to measure the line integral of some plasma parameters, such as the electron density and the magnetic field, and constrain, analyse and validate the equilibrium models. Despite the strong link between the plasma properties and light polarisation propagation, the interpretation of plasma polarimetry remains complex and sometimes uncertain. The type 1 approximation is the most common hypothesis used to link the polarisation effects, such as the Faraday rotation and the Cotton Mouton phase shift, with the plasma properties (electron density and magnetic fields). However, this approximation is valid only in specific conditions, which depend on both the plasma configuration and initial polarisation of the electromagnetic wave. Moreover, the uncertainty propagations of these measurements are affected by regions where their values are too high to guarantee accurate values, making the measurement unsuitable. The need to measure an unsteady physical environment, which goes from the condition of no plasma to the flat top of the discharge, makes the setting of the measurement more constrained. The aim of this work is to analyse, both analytically and numerically, the behaviours of the interpretability and uncertainty propagation of polarisation measurements, in order to give a detailed and the most general as possible description of these issues, ensuring an easier, more performant and reliable understanding and design of plasma polarimetry. The results will show that an input linear polarisation around 45° degree is the most suitable for plasma polarimetry when type 1 approximation is adopted and which the choice of the laser wavelength governs the performances of the polarimeter.