We note that compared with the FSR and group indices calculated during the boundary condition simulations we find the results to be below the expected value. This could be for a variety of reasons (and probably a combination of them). First, from the boundary condition calculations done in simulation, note that increasing the width decreases the group index. Thus the fabrication of the waveguides could have been larger than what we expected. Second, decreasing the height also decreased the group index. In prior courses and fabrication runs it has been noted that the height is almost never at 220nm, instead it's lower, around 210nm. The boundary conditions we performed were in the range of 215.3nm to 223.1nm. Thus the actual fabrication could be outside of the range of these boundary conditions. Third, if the delta length of the MZI is skewed slightly, say from 399 to 400 microns, this could also have an effect on the resultant curve fitting. For example, one of the readings for the 100 micron MZI gave an estimated group index of 4.16. Decreasing the MZI's delta L to 99 gave an estimated group index of 4.20! Fourth, a poor MZI fitting could also contribute to the lower values, although only marginally. If we were to hazard a guess, we believe it could be due to a combination of the height and width being slightly outside of our variances as well as a slight decrement of the length of the MZI.