Analysis and presentation of intracellular measurements obtained with ion-selective microelectrodes.

We have considered the manner in which data obtained with ion-selective electrodes should be evaluated. The potential difference recorded by such electrodes, with respect to a stable reference, is converted to a concentration by a non-linear transformation--the Nernst or Nikolsky equation. The mean and standard deviation of such estimations of concentration are then usually presented, which assumes that the latter variable is normally distributed. If, however, the recorded potential difference (PD) is the normally distributed variable and the mean value calculated, then a different value of mean concentration will be obtained. We show here that the recorded PD is indeed the normally distributed variable using data from a variety of ion-selective electrode measurements and conclude that the mean values of quoted ion concentrations have been overestimated by 6-43%.