Key points are not available for this paper at this time.
I have tried to talk about three areas in which I think newer techniques of data analysis will come into, or be developed in, geophysics which will advance our knowledge about the plasma-like, gaseous, liquid, and solid earth: (i) developments along the line of the concepts of the spectrum, with emphasis both on the use of cross-spectral methods of studying relationship and on bispectrum and higher methods of studying nonlinearities (including modulations) and frequency interactions; (ii) wise choice of modes of expression for analysis; (iii) the use of methods appropriate and effective in situations where distributions of errors and fluctuations do not have exactly the magic bell-shaped distribution.
John W. Tukey (Fri,) studied this question.