Polarization-resolved second harmonic generation (P-SHG) microscopy is commonly used to analyze the second-order susceptibility, χ(2), tensor, which enables the calculation of the molecular structure of harmonophores. However, despite extensive research on type I collagen, the measured χ(2) ratios vary considerably among published values, which raises the question of whether P-SHG imaging is universally applicable to all tissues containing harmonophores. In this work, we propose that the deviation of χ(2) ratios is primarily due to ignoring the molecular tilt angle and chirality in image analysis. To confirm our hypothesis, we present an analytical model based on C6 symmetry that takes into account these two factors. We also introduce an imaging scheme that splits SHG into X- and Y-polarized image components for χ(2) tensor analysis. Our approach effectively improves the precision of determining χ(2) ratios, depending on how much the two factors affect the P-SHG signals.