Showing papers on "Fisher transformation published in 1997"
102 citations
TL;DR: In this article, an interpretation of the Pearson productmoment correlation coefficient is presented that describes the coefficient in terms of the proportion of matches in an idealized data set, which is then expressed as the result of the definition of the PPI correlation.
Abstract: An interpretation of the correlation coefficient is presented that describes the coefficient in terms of the proportion of matches in an idealized data set. This characteristic is then expressed as the result of the definition of the Pearson product-moment correlation. An expression is derived to relate the correlation coefficient to the proportion of times a second variable falls within a set range of the first. Both a simulation and a real data example are presented to show how this would be reflected in data.
54 citations
TL;DR: In this article, robust estimation of the correlation coefficient of a bivariate normal distribution is considered in the case of a contamination scheme, and a number of conventional robust estimates are studied, and some new estimates are proposed.
Abstract: Robust estimation of the correlation coefficient of a bivariate normal distribution is considered in the case of a contamination scheme. A number of conventional robust estimates are studied, and some new estimates are proposed. Their properties are examined on finite samples and in asymptotics with the use of Monte-Carlo and the influence functions techniques correspondingly. It is shown that one of the proposed estimates called a median correlation coefficient has high robustness properties.
34 citations
TL;DR: The authors showed that unless Y = kX, k > 0, the coefficient of determination (ϱ2) decreases monotonically as the differencing interval increases, approaching zero in the limit.
25 citations
Posted Content•
9 citations
TL;DR: The maximum likelihood estimate and the restricted or residual maximum likelihood estimates for a common intraclass correlation coefficient among several bivariate normal distributions when some observations on either of the variables are missing are given as the solutions of polynomial equations as discussed by the authors.
Abstract: SYNOPTIC ABSTRACTThe maximum likelihood estimate and the restricted or residual maximum likelihood estimate are considered for a common intraclass correlation coefficient among several bivariate normal distributions when some observations on either of the variables are missing The estimates are given as the solutions of polynomial equations Asymptotic variances of both estimates are obtained from the corresponding information matrices The variance stabilizing transformation, which can be used to perform hypothesis tests and construct a confidence interval for ρ, is derived
4 citations
01 Nov 1997
TL;DR: In this paper, the existence of the best monotonic lower confidence limit for unknown parameter of distribution is proved for the multiple correlation coefficient in the case of the normal law. But the existence is not proved for any other parameter.
Abstract: We introduce a notion of the best monotonic lower confidence limit for unknown parameter of distribution. Its existence is proved for the multiple correlation coefficient in the case of the normal law. An effective method of computing such a limit is described.
1 citations