Modified large sample confidence intervals for Poisson distributions: Ratio, weighted average, and product of means

2016 ◽  
Vol 45 (1) ◽  
pp. 83-97 ◽  
Author(s):  
K. Krishnamoorthy ◽  
Jie Peng ◽  
Dan Zhang
Keyword(s):  
Confidence Intervals ◽  
Weighted Average ◽  
Large Sample ◽  
Poisson Distributions ◽  
Modified Large Sample

scholarly journals Brief communication "Improving the actual coverage of subsampling confidence intervals in atmospheric time series analysis"

2012 ◽  
Vol 19 (5) ◽  
pp. 473-477
Author(s):  
A. Gluhovsky ◽  
T. Nielsen
Keyword(s):  
Time Series ◽  
Time Series Analysis ◽  
Confidence Intervals ◽  
Coverage Probability ◽  
Resampling Methods ◽  
Large Sample ◽  
Series Analysis

Abstract. In atmospheric time series analysis, where only one record is typically available, subsampling (which works under the weakest assumptions among resampling methods), is especially useful. In particular, it yields large-sample confidence intervals of asymptotically correct coverage probability. Atmospheric records, however, are often not long enough, causing a substandard coverage of subsampling confidence intervals. In the paper, the subsampling methodology is extended to become more applicable in such practically important cases.

An Empirical Comparison of the Large Sample Approximations for the Coefficient of Concordance and Weighted Average Tau

1978 ◽  
Vol 3 (3) ◽  
pp. 253-264
Author(s):  
Carol Lee Stamm
Keyword(s):  
Weighted Average ◽  
Simple Approximation ◽  
Data Sets ◽  
Empirical Comparison ◽  
Large Sample ◽  
Empirical Distributions ◽  
Coefficient Of Concordance

Two rather similar statistics, the coefficient of concordance (W) and weighted average tau (Wa), have been suggested for estimating the communality between rankings. Two large sample tests have been suggested for each statistic when the values of m (the number of judges) and n (the number of objects) are larger than tabled values. The purpose of this study was to determine empirically which of the large sample approximations for the statistic, W or Wa, was more appropriate for selected values of m and n. The procedures for the investigation involved manipulating m × n rankings using generated data sets. The results of the comparisons of the distributions of the large sample approximations for W and Wa with their empirical distributions indicated that neither approximation was better for all sets of value m and n that were examined. However, the results suggest that the computationally simple approximation, χ2r for W and χ2n for Wa, are appropriate for values of m and n chat are smaller than those recommended previously.

Sign in / Sign up

Export Citation Format

Share Document