The types of measurand parameter estimators derived from samples of measured data taken from a sym. trapezoidal population were briefly reviewed (9 refs.). A non-std. approach to find ests. of the non-Gaussian distributions parameters based on the unconventional method for maximizing the stochastic polynomials by using a moment-cumulant description of random variables was proposed. The method was recommended to use for detg. estd. values of the std. deviation and uncertainties of measurand when distribution of the random errors population is a priori unknown and first few cumulants have to be found from the sample data. The method is particularly useful in assessing mixts. and mixing efficiency.