A Mach-Zehnder interferometer (MZI) is used to decouple the electro-optic (EO) and piezoelectric tensor components for a poled polymer film. In the past those using the MZI method failed to take into account the piezoelectric contribution in the polymer which can lead to erroneous EO coefficient data. The typical poled sample of polymer sandwiched between ITO glass and gold that was developed for the popular Teng-Man reflection ellipsometry method is used, providing for easy comparison with that method. The sample serves as a mirror in one arm of the interferometer with the gold side facing the beam for measuring the piezoelectric modulation and the glass side facing the beam to measure the coupled piezoelectric and EO modulation. Optical biasing in the reference arm allows for the baseline and modulated contrast of the system to be measured from which the tensor components are calculated. This method has the advantage over the reflection ellipsometry method of allowing for the independent determination of the Pockel's coefficients r13 and r33 and the piezoelectric coefficient d33. The r33 value of a guest host polymer that consists of AJLZ53/amorphous polycarbonate (APC) was found to be 122.7 pm/V and 123.0 pm/V for the MZI and reflection ellipsometry method respectively. The r33 data fits well to the dispersion of the second order susceptibility tensor χ333(2) based on the two-level model approximation. Measurements were done from 100 Hz to 100 kHz with the results showing that at higher frequencies the mechanical effects in the sample are negligible and modulation is almost entirely due to the EO effect, as expected.