Digital power spectral and wave polarization analysis are powerful techniques for studying ULF waves in the earth's magnetosphere. Four different techniques for using the spectral matrix to perform such an analysis have been presented in the literature. Three of these techniques are similar in that they require transformation of the spectral matrix to the principal axis system prior to performing the polarization analysis. The differences in the three techniques lie in the manner in which determine this transformation. A comparative study of these three techniques using both simulated and real data has shown them to be approximately equal in quality of performance. The fourth technique does not require transformation of the spectral matrix. Rather, it uses the measured spectral matrix and state vectors for a desired wave type to design a polarization detector function in the frequency domain. The design of various detector functions and their application to both simulated and real data will be presented.