Simple and accurate method to evaluate type a standard and expanded uncertainties of measurement for the Laplace distributed observations
Year: 2024
Type: article
Abstract: The article proposes simple and accurate evaluation of the standard and expanded uncertainties of the Laplace population median. The uncertainty evaluation based on the approximation of the population median distribution for n observations by the distribution of the sample median for n - 2 observations. Comparisons of the results obtained by proposed approximation with exact results shown that when 6 ≤ n ≤ 10 the differences between approximated and exact standard uncertainties are lees then 1% (0.56%), i.e. negligibly small and for a confidence level p = 0.90 and 0.95 the approximated value of expanded uncertainty differ from exact value less than 1%, and for p= 0.99 difference does not exceed ≈ 5%. When n ≥ 11 the Monte Carlo simulation results showed that the proposed method for determining both the standard and the expanded uncertainty of population median do not exceed ±1%, i.e. is very much precise.
Author: Mykhaylo Dorozhovets
Language: English
Subfield: Statistics, Probability and Uncertainty
Field: Decision Sciences
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W4407991529