Statistical Analysis of Experimental Data: A Fallacy in Judging Non‐Normal Distributions by a Logarithmic or Functional Probability Paper Plot

Abstract

PROBABILITY paper was originally used to represent the summation curve of a Gaussian distribution as a straight line. Gibrat, Divisia in France and Daeves and Beckel in Germany have shown, that the summation curve of most skew distributions may also be represented as a straight line, if the argument x is traced as log (±x±a), where a is an arbitrarily chosen constant, which may, but need not, represent a physical condition.

Citation

Rossow, E. (1949), "Statistical Analysis of Experimental Data: A Fallacy in Judging Non‐Normal Distributions by a Logarithmic or Functional Probability Paper Plot", Aircraft Engineering and Aerospace Technology, Vol. 21 No. 12, pp. 378-381. https://doi.org/10.1108/eb031837