The problem of estimating the variance of the Wilcoxon-Mann-Whitney statistic is considered. Minimum variance unbiased estimator which is a U-statistic, bootstrap estimator, an estimator proposed by Fligner & Policello (1981) and jackknife estimator are considered. We also consider induced estimators of the standard deviation and interval estimates of the expectation of the Wilcoxon-Mann-Whitney statistic. By computer simulations, we conclude that bootstrap estimator is efficient for both the variance and the standard deviation in the sense of mean squared error. This is the same conclusion with Sakamoto & Shirahata (1992) where the estimating problem of variance of one-sample U-statistics is considered. It is, in addition to the above, found that the mean squared error of minimum variance unbiased estimator is small and the bias of the induced estimator is also small for the standard deviation. Fligner & Policello estimator gives accurate confidence coefficients for many cases and the jackknife estimator stands between the former two estimators and Filgner & Policello estimator.

View full abstract

In cases where the number of predetermined variables exceeds the sample size or strong multicollinearity occurs among excluded predetermined variables, the limited information maximum likelihood and the least variance ratio cannot be used to estimate a simultaneous equation system. We propose a method in which some effectively-influential principal components of excluded predetermined variables are employed as instrumental variables in place of the excluded predetermined variables. Then we demonstrate the estimation of a Cobb-Douglas agricutural production function with the proposed method.

View full abstract

The statistical software SAM II which is a revised version of SAM is designed to make sensitivity analysis in descriptive multivariate analysis. This program has the unified formulation of sensitivity analysis for any multivariate methods formulated as eigenvalue problems by using generalized singular value decomposition. In this program there are three steps; (1) pre-processing, (2) original analysis and (3) sensitivity analysis. By this revision, the number of multivariate methods to be treated are increased, the displays of the influences of individuals on the results of analysis are improved and the influence of not only a single individual but also multiple individuals can be observed.

View full abstract

When a pharmaceutical manufacturer develops a drug, various documents are produced in the process. For example: Statistical material of clinical trials, summary reports, and material for the medicine approval application submitted to the Ministry of Health and Welfare. In these documents, two major types of reports are used. One is case report tabulation where the clinical data are integrated, and another is statistical reporting where various statistics and results of tests are described. The system which is introduced here can produce statistical reports necessary during the process of medical drug development.

View full abstract

An aptitude test was carried out on 265 junior high school students. The obtained data were analyzed by means of principal component analysis (PCA) and principal factor analysis (PFA) to explore the number and rough sketch of factors and then by mean of LISREL and its sensitivity analysis to investigate the structure of the test more precisely. It was found that the test was composed of three factors, i. e., the factors of coding and searching ability, symbolic processing ability and logical thinking ability, and that those results were not unstable for small changes of the data. It was also found that LISREL and its sensitivity analysis were useful for the analysis of this kind of data.

View full abstract

An estimator, called initial estimator, is proposed for the complementary log-log model when the binary response data are observed. It is shown that this estimation can be used as an estimator of the true parameter and that this estimator can be adopted as an initial value when computing the value of maximum likelihood estimate. It is also shown that in a simulation study, our proposed initial estimator has smaller bias than that of maximum likelihood estimator for in a case when the size of sample is small.

View full abstract

Investment in stocks has a great deal of danger compared to investment in time deposits. Naturally, it becomes important for individual investors to know and understand the short-term fluctuation of the market price.
This study examines the possibility of obtaining profit of investment by knowing the fluctuation of the stock price with regression analysis, even when there is a circumstance which makes a fluctuation of the market price downward. We still have opportunities to invest on stocks when the stock price is in this falling phase.
There has been an opinion that investors will always gain a margin of profit when the stock price is in the rising phase, yet we only know later that the price was rising or falling at a particular moment.

View full abstract