Robustness of the two - sample t - test under violations of the homogeneity of variance assumption,part ii

In a previous paper, Posten, Yeh and Owen (1982), the robustness of the type I error for the two tailed two sample t - test was studied under departures from the assumption of equal variances. The level of robustness of this test was then quantified under the concept of regions of robustness. These results are extended here to the one - tailed test for the same problem. The high level of robustness for equal or nearly equal sample sizes observed in the previous study is again documented quantitatively.     

The use of Fourier series in the evaluation of probability distribution functions

In statistics, Fourier series have been used extensively in such areas as time series and stochastic processes. These series; however, to a large degree have been neglected with regard to their use in statistical distribution theory. This omission appears quite striking when one considers that, after the elementary functions, the trigonometric functions are the most important functions in applied mathematics. In this paper a procedure is developed for utilizing Fourier series to represent distribution functions of finite range random variables as Fourier series with coefficients easily expressible (using Chebyshev polynomials) In terms of the moments of the distribution. This method allows the evaluation of probabilities for a wide class of distributions. It is applied to the     

Conference Report Making Statistics More Effective in Schools of Business

We propose semiparametric procedures for estimation and testing of base-independent equivalence scales. The partial linear index specification permits simultaneous estimation across multiple household types and multiple goods and also the incorporation of continuous and discrete household attributes. Furthermore, asymptotic properties of estimated equivalence scales and tests of base independence are readily obtained. The efficiency gains from the proposed models and estimators are particularly helpful for developing country data where there is often much greater variation in household size and composition. We apply the techniques to South African data and find the results to be broadly consistent with base independence.     

一种理解西方人类学发展的历史视角

仔细考察了16世纪和19世纪具体历史事件的相关性.在这些具体的历史事件中,一种对异族异群体的体验产生出了一种人类学的视角.如此一种视角在辩证意叉上是受19世纪中期日益变得清晰的生物学意义上的种族观点所影响的.这样一种文化的视角在争议中渐成支配,并在解释和评论哲学和心理学议题的兼容性方面取得其基础性的地位.     

Monetary policy in the New-Keynesian model: An application to the Euro Area

This paper analyzes monetary policy in a stylized New-Keynesian model. A number of issues are focused upon: (i) optimal monetary policy under commitment or discretion versus ad-hoc monetary policy based on simple rules, (ii) the effects of fiscal policies and foreign variables on monetary policy, (iii) the effects of fiscal deficit and interest rate smoothing objectives and the role of forward-backward linkages in the model. The model is estimated for the Euro Area. Using simulations of the estimated model, it is analyzed how these aspects might affect monetary policy of the ECB and macroeconomic fluctuations in the Euro Area.     

Accent on Teaching Materials: Notices and Reviews of Important Teaching Materials

In contrast to the analysis of variance of fully fixed or fully random component models, the analysis of variance of mixed models is fraught with potential pitfalls. It is fortunate that there are simple rules for the correct analysis of balanced data; in the case of unbalanced data there are no simple results. The potential pitfalls in the path of a correct analysis are well-known. Despite this, some computer packages still report incorrect results for the balanced model and some textbooks gloss over or ignore some of these pitfalls.     

Die Schere öffnet sich

Recent decades have witnessed some extraordinary increases of income richness in some countries. Even though many consider Germany to be different, recent research has found a strong increase of rich income households in Germany, as well. However, Germany’s data infrastructure on rich households is comparatively weak so that little is known on the causes of higher richness headcounts. This article discusses research and possible causes of the recent increase of household richness using a decomposition of unconditional quantile regressions and data from the German Socio-economic Panel (GSOEP). Comparing most recent data and the early 1990s after reunification, the most striking result is that the upper half of the German income distribution was polarized most by taxes and social security contributions which reinforce labor market related developments.