A continuous information theoretic approach to the analysis of cutaneous receptor neurons.

Information theoretic and statistical techniques for determining the number of discernible levels in cutaneous receptor neurons are reviewed. Reasons for the large variance in these results are discussed. A new continuous information theoretic analysis technique is presented that overcomes many of the problems in the other methods of analysis discussed. Comparison of this new method of analysis with a statistical technique developed by Schreiner et al. (1978) clearly shows some of the misconceptions that are associated with statistical analysis techniques, and why these problems cannot arise in the new information theoretic technique discussed here.