ABSTRACT
In part to deal with goodness of fit follow-up issues, during the 1950s and ’60s the histogram was promoted from its job as a check against some estimator like f ( x ) = [x — μ ]/σ ) /σ , based on estimators, μ and σ, and began to be assigned other more demanding responsibilities. One such responsibility was as a substitute for the curve f ( x ) that appears within an expression that defines a particular curve property.
