Materials scientists and engineers use optical tweezer-based microrheology measurements to probe the viscoelastic properties of complex microstructured materials such as gels and filament networks with various structural feature sizes at multiple length scales by trapping and tracking beads of various sizes suspended within the materials. Current procedures for analyzing these data assume that the detection of the bead within the trap and the force of the trap on the bead are linear. These assumptions inherently rely on an approximation that the bead diameter is similar to the beam width at the diffraction-limited focal point of the trapping laser, but less dense materials have larger structural features and require larger diameter beads. We discovered significant deviations from the linear assumption in the detection of beads that are 5-10-fold larger than the laser focal point. Additionally, we found nonlinearities in the force-displacement curves of the trap when we applied known constant forces to 5 µm trapped beads. We customized MATLAB toolboxes to theoretically model and validate our experimental data. Using both the calculation and experimental results, we found that that the magnitude and direction of the slope and linear range of the detector-response graphs are strong functions of bead size for large beads that can be modeled as third-order polynomials. Further, the force-displacement modeling and data that we collected suggest that the trap has a non-linear effective spring constant for large beads. By quantifying these observations with the analysis of large bead optical tweezer experiments, we better understand their interaction with the trap and how to account for these nonlinearities properly. Ultimately, our results will allow researchers to analyze sparse material matrices, use larger beads, and accurately perform calculations with optical tweezer-based microrheology.