Search

ABSTRACT The spread between the lending and the deposit rates plays a vital role in accelerating financial sector development of any country. This paper examines the determinants of the interest [...]

This paper tests three hypotheses about the origins of the conflict which began in 2014 in Ukraine, using the 1995-2011 World Values Surveys. First, a hypothesis that the economic situation of young fighting-age men in Eastern Ukraine worsened relative to that of young men in Russia during 1995-2011 is examined. Second, a hypothesis that the political views of respondents in Eastern Ukraine became more like those of people in neighbouring Russian regions is investigated. Third, a hypothesis that people in Eastern Ukraine became relatively more disillusioned with the quality of their national institutions during this period is tested. None of these three hypotheses is much supported by the data. Comparative Economic Studies (2016) 58, 606-637. doi:10.1057/s41294-0160007-z; published online 30 November 2016 Keywords: Russia, Ukraine, World Values Surveys (WVS), opinions JEL Classification: H00, J2, P3, P5, INTRODUCTION This paper examines whether or not a socioeconomic and political divide between Russian and Ukrainian-speaking regions of Ukraine emerged between 1995 and 2011, causing residents of Russian-speaking regions to [...]

Special care must be exercised in the interpretation of multiple statistical hypothesis tests--for example, when each of many tests corresponds to a different location. Correctly interpreting results of multiple simultaneous [...]

Most of the study of causal relationship of GDP or growth with Trade and Foreign Direct Investment in the context of India rely on cross country comparisons, or simple correlations/regressions disregarding nonstationarity properties, or at most on VAR or VECM or Johansen-Juselius cointegration technique of testing for Granger causality. The potential biases, pitfalls, influence of nuisance parameters and asymptotic unreliability of these techniques have been well documented in the literature and also briefly mentioned in this paper. Therefore, this paper employs the more recent and robust Toda-Yamamoto-Dolado-Lutkephol augmented VAR(p) technique for testing Granger causality among these three time series. This technique has been shown to provide more robust and asymptotically reliable results under wide variety of situations regarding the cointegration relationships among the time series. This study also focuses on the post liberalization period and clearly shows that the post-liberalization period significantly differs from the pre-liberalization period in the GDP-Export-FDI nexus. We find strong support for Export-led and Foreign Direct Investment led growth hypotheses only in the post liberalization period. JEL Classifications: O4 C22 Keywords: Export-Led-Growth, FDI, Applied Econometrics, Granger Causality, INTRODUCTION In every respect India is an interesting and increasingly important case for study of relationship between growth, Trade and Foreign direct Investment. India is the second largest country in [...]

Likelihood ratio (LR) tests for association and for interaction are examined for three-way contingency tables, particularly the widely used 2 x 2 x K table. Mutual information identities are used to characterize the information decomposition and the logical relationship between the omnibus LR test for conditional independence across K strata and its two independent components, LR tests for interaction and for uniform association. The latter two tests are logically connected to formulating a natural two-step test for conditional independence. The proposed two-step test with reduced nominal levels is suggested instead of the Breslow-Day test and the Cochran-Mantel-Haenszel test. This yields efficient interval estimation for both the interaction parameter and the common odds ratio compared with using the Mantel-Haenszel estimate. This allows the development of power analysis for testing general hypotheses of varied interactions, using an invariant Pythagorean law of relative entropy. KEY WORDS: Breslow-Day test; Cochran Mantel-Haenzel test; Mutual information; Pearson test; Three-way interaction.

In this paper we develop multivariate outlier tests based on the high-breakdown Minimum Covariance Determinant estimator. The rules that we propose have good performance under the null hypothesis of no outliers in the data and also appreciable power properties for the purpose of individual outlier detection. This achievement is made possible by two orders of improvement over the currently available methodology. First, we suggest an approximation to the exact distribution of robust distances from which cut-off values can be obtained even in small samples. Our thresholds are accurate, simple lo implement and result in more powerful outlier identification rules than those obtained by calibrating the asymptotic distribution of distances. The second power improvement comes from the addition of a new iteration step after one-step reweighting of the estimator. The proposed methodology is motivated by asymptotic distributional results. Its finite sample performance is evaluated through simulations and compared to that of available multivariate outlier tests. KEY WORDS: Minimum covariance determinant estimator; Multiple outliers; Reweighting; Robust distance; Size and power.

We develop a testing methodology that can be used to predict the performance of e-mail marketing campaigns in real time. We propose a split-hazard model that makes use of a [...]

Functional magnetic resonance imaging (fMRI) data that are corrupted by temporally colored noise are generally preprocessed (i.e., prewhitened or precolored) prior to functional activation detection. In this paper, we propose likelihood-based hypothesis tests that account for colored noise directly within the framework of functional activation detection. Three likelihood-based tests are proposed: the generalized likelihood ratio (GLR) test, the Wald test, and the Rao test. The fMRI time series is modeled as a linear regression model, where one regressor describes the task-related hemodynamic response, one regressor accounts for a constant baseline and one regressor describes potential drift. The temporal correlation structure of the noise is modeled as an autoregressive (AR) model. The order of the AR model is determined from practical null data sets using Akaike's information criterion (with penalty factor 3) as order selection criterion. The tests proposed are based on exact expressions for the likelihood function of the data. Using Monte Carlo simulation experiments, the performance of the proposed tests is evaluated in terms of detection rate and false alarm rate properties and compared to the current general linear model (GLM) test, which estimates the coloring of the noise in a separate step. Results show that theoretical asymptotic distributions of the GLM, GLR, and Wald test statistics cannot be reliably used for computing thresholds for activation detection from finite length time series. Furthermore, it is shown that, for a fixed false alarm rate, the detection rate of the proposed GLR test statistic is slightly, but (statistically) significantly improved compared to that of the common GLM-based tests. Finally, simulations results reveal that all tests considered show seriously inferior performance if the order of the AR model is not chosen sufficiently high to give an adequate description of the correlation structure of the noise, whereas the effects of (slightly) overmodeling are observed to be less harmful. Index Terms--Functional magnetic resonance imaging (fMRI), generalized likelihood ratio test, Rao test, statistical parametric maps, time series analysis, Wald test.

Groupwise registration of a set of shapes represented by unlabeled point sets is a challenging problem since, usually, this involves solving for point correspondence in a nonrigid motion setting. In this paper, we propose a novel and robust algorithm that is capable of simultaneously computing the mean shape, represented by a probability density function, from multiple unlabeled point sets (represented by finite-mixture models), and registering them nonrigidly to this emerging mean shape. This algorithm avoids the correspondence problem by minimizing the Jensen-Shannon (JS) divergence between the point sets represented as finite mixtures of Gaussian densities. We motivate the use of the JS divergence by pointing out its close relationship to hypothesis testing. Essentially, minimizing the JS divergence is asymptotically equivalent to maximizing the likelihood ratio formed from a probability density of the pooled point sets and the product of the probability densities of the individual point sets. We derive the analytic gradient of the cost function, namely, the JS-divergence, in order to efficiently achieve the optimal solution. The cost function is fully symmetric, with no bias toward any of the given shapes to be registered and whose mean is being sought. A by-product of the registration process is a probabilistic atlas, which is defined as the convex combination of the probability densities of the input point sets being aligned. Our algorithm can be especially useful for creating atlases of various shapes present in images and for simultaneously (rigidly or nonrigidly) registering 3D range data sets (in vision and graphics applications), without having to establish any correspondence. We present experimental results on nonrigidly registering 2D and 3D real and synthetic data (point sets). Index Terms--Groupwise point set registration, finite-mixture model, Jensen-Shannon divergence, hypothesis testing, thin-plate spline.

Even in a single-tissue type cancer is often a collection of different diseases, each with its own genetic mechanism. Consequently, a gene may be expressed in some but not all of the tissues in a sample. Differentially expressed genes are commonly detected by methods that test for a shift in location that ignore the possibility of heterogeneous expression. This article proposes a two-sample test statistic designed to detect shifts that occur in only a part of the sample (partial shifts). The statistic is based on the mixing proportion in a nonparametric mixture and minimizes a weighted distance function. The test is shown to be asymptotically distribution free and consistent, and an efficient permutation-based algorithm for estimating the p value is discussed. A simulation study shows that the test is indeed more powerful than the two-sample t test and the Cramer von Mises test for detecting partial shifts and is competitive for whole-sample shifts. The use of the test is illustrated on real-life cancer datasets, where the test is able to find genes with clear heterogeneous expression associated with reported subtypes of the cancer. KEY WORDS: Asymptotics: Consistency: Differential gene expression; Non parametric mixture; Permutation test; Weighted minimum distance.

A point selection scheme based on multiple hypothesis testing is proposed for blind separation of nonstationary signals. Results suggest the achievement of good blind source separation (BSS) performance at lower signal-to-noise ratio (SNR).

We discuss methodology for controlling the false discovery rate (FDR) in complex large-scale studies that involve testing multiple families of hypotheses; the tested hypotheses are arranged in a tree of disjoint subfamilies, and the subfamilies of hypotheses are hierarchically tested by the Benjamini and Hochberg FDR-controlling (BH) procedure. We derive an approximation for the multiple family FDR for independently distributed test statistics: q, the level at which the BH procedure is applied, times the number of families tested plus the number of discoveries, divided by the number of discoveries plus 1. We provide a universal bound for the FDR of the discoveries in the new hierarchical testing approach, 2 x 1.44 x q, and demonstrate in simulations that when the data has an hierarchical structure the new testing approach can be considerably more powerful than the BH procedure. KEY WORDS: False-discovery rate; Hierarchical testing.

We propose a new method to uniformly improve cutoff tests (as, e.g., the Bonferroni test) by replacing them with a sequential testing procedure. The proposed procedure provides a uniform improvement--that is, it rejects the null hypothesis at least in all cases where the original procedure does, has level [alpha], and never requires a sample size larger than that of the original procedure. Information on the often-unknown joint distribution of the test statistics across the hypotheses is not required. We discuss the applications to one- and two-sided tests for the intersection of multiple hypotheses, as well as to tests of elementary hypotheses. For the wide class of asymptotically linear test statistics and under unknown nuisance parameters, we propose an asymptotic approximation to our improved procedure and show that it exhausts the nominal type I error rate asymptotically. We present an application in a clinical study with multiple endpoints where the null hypothesis of no effect is tested against the alternative that at least one endpoint shows an effect, and give a real data example. KEY WORDS: Bonferroni test; Cutoff tests; Multiple endpoints; Multiple testing; Sequential test.

We analyze experimental survey data, with a random split into respondents who get an open-ended question on the amount of total family consumption (with follow-up unfolding brackets of the form 'Is consumption $X or more?' for those who answer 'don't know' or 'refuse') and respondents who are immediately directed to unfolding brackets. In both cases, the entry point of the unfolding bracket sequence is randomized. Allowing for any type of selection into answering the open-ended or bracket questions, a nonparametric test is developed for errors in the answers to the first bracket question that are different from the usual reporting errors that will also affect open-ended answers. Two types of errors are considered explicitly: anchoring and yea-saying. Data are collected in the 1995 wave of the Assets and Health Dynamics survey, which is representative of the population in the United States that is 70 years and older. We reject the joint hypothesis of no anchoring and no yea-saying. Once yea-saying is taken into account, we find no evidence of anchoring at the entry point. KEY WORDS: Acquiescence bias; Item nonresponse; Unfolding brackets.

A report on a low-cost, compositional approach based on the use of the four statistical moments to characterize the failure time distributions of the constituent components, subsystems and top-level systems using Pearson distributions as an intermediate analytical vehicle gave good accuracy at very modest cost is presented.

Using procedures identical to a study published 10 years earlier, parents of 49 children with cancer (48 mothers, 33 fathers) and 49 healthy classmates (49 mothers, 29 fathers) completed measures [...]

Multiple testing methods to control the false discovery rate, the expected proportion of falsely rejected null hypotheses among all rejections, have received much attention. It can be valuable to control not the mean of this false discovery proportion (FDP), but rather the probability that the FDP exceeds a specified bound. In this article we construct a general class of methods for exceedance control of FDP based on inverting tests of uniformity. The method also produces a confidence envelope for the FDP as a function of rejection threshold. We discuss how to select a procedure with good power. KEY WORDS: False discovery rate; Multiple testing; p value.

Recently the Oaxaca decomposition for wage differentials has been extended to study the differences in the first moment. For example, this type of decomposition can be used to study differences in labor market participation rates by race, where the participation equation may be estimated by logit or probit. This paper develops a simple test for the significance of characteristics and coefficients effects in a decomposition analysis of differences in the first moment at aggregate and individual variable levels. In order to derive the test statistics, the asymptotic variances of the characteristics and coefficients effects at both aggregate and individual variable levels are calculated using the well-known delta method. The significance test for the Oaxaca decomposition is incorporated as a special case of this proposed test. JEL codes: J70, C12, C20. Keywords: Decomposition analysis, significance test, characteristics effect, coefficients effect

There is agreement among leading theorists and econometricians on two main points: first, that economic significance usually has nothing to do with statistical significance and second, that a majority of economists do not explore economic significance in their research. This agreement should by itself lead to change in research practice. It is urged that economic significance is what science and citizens want and need.

The problem of hypothesis testing about proportions in two finite populations is addressed. The usual test based on the normal approximation (Z test) and a test based on estimated p values (E test) are considered. The exact properties of the tests are evaluated numerically. Numerical studies indicate that the E test is very satisfactory even for small samples and can be recommended for practical use. Power computation for a given level and sample size is also outlined for both tests. The testing methods are illustrated using an example. KEY WORDS: Binomial distribution; Exact power; Exact size; Hypergeometric distribution; Sample size calculation; Standardized difference., 1. INTRODUCTION There has been a continuous interest in developing small sample exact inferential procedures for discrete distributions. The growing interest in this area is partly because modern computing technology [...]

This paper addresses the problem of detecting small, moving, low amplitude objects in image sequences that also contain moving nuisance objects and background noise. We formulate this problem in the context of a hypothesis testing procedure on individual pixel temporal profiles, leading to a computationally efficient statistical test. The technique assumes we have reasonable deterministic and statistical models for the temporal behavior of the background noise, target, and clutter, on a single pixel basis. Based on these models we develop a generalized likelihood ratio test (GLRT) and perfect measurement performance, analysis, and present the resulting decision rule. We also propose a parameter estimation technique and compare its performance to the Cramer Rao bound (CRB). We demonstrate the effectiveness of the technique by applying the resulting algorithm to real world infrared (IR) image sequences confining targets of opportunity. The approach could also be applicable to other image sequence processing scenarios, using acquisition systems besides IR imaging, such as detection of small moving objects or structures in a biomedical or biological imaging scenario, or the detection of satellites, meteors or other celestial bodies in night sky imagery acquired using a telescope.

The author critiques the use of null hypothesis significance testing, asserting that conditions in which the method is useful rarely arise in wildlife science. He instead supports the use of replication to produce consistent research results.

Business failure prediction is one of the most essential problems in the field of financial management. The research on developing quantitative business failure prediction models has been focused on building discriminant models to distinguish among failed and non-failed firms. Several researchers in this field have proposed multivariate statistical discrimination techniques. This paper explores the applicability of multicriteria analysis to predict business failure. Four preference disaggregation methods, namely the UTADIS method and three of its variants, are compared to three well-known multivariate statistical and econometric techniques, namely discriminant analysis, logit and probit analyses. A basic (learning) sample and a holdout (testing) sample are used to perform the comparison. Through this comparison, the relative performance of all the aforementioned methods is investigated regarding their discriminating and predicting ability.

This article presents a method to ascertain exact loci in a genome and its bearing in identifying the gene that affects quantitative traits. The author proposes the term 'suggestive linkage' for an objective comparison of results across species in qunatitative trait loci analyses.

A study of generality of association between fluctuating asymmetry (FA), random departure from perfect bilateral symmetry, and heterozygosity (H) has been carried out using meta-analysis by converting test statistics for relationship into a common effect size, Pearson's product-moment correlation coefficient. A database with 41 studies and 118 individual samples was used. It has been hypothesized that there is an inverse correlation, and the claim is widely repeated, but some studies dispute it. Weak association at the most was found.

It is shown that partial classification, which allows for indecision in certain regions of the data space, can increase a benefit function, defined as the difference between the probabilities of correct and incorrect decisions, joint with the event that a decision is made. This is particularly true for small data samples, which may cause a large deviation of the estimated separation surface from the intersection surface between the corresponding probability density functions. Employing a particular density estimation method, an indecision domain is naturally defined by a single parameter, whose optimal size, maximizing the benefit function, is derived from the data. The benefit function is shown to translate into profit in stock trading. Employing medical and economic data, it is shown that partial classification produces, on average, higher benefit values than full classification, assigning each new object to a class, and that the marginal benefit of partial classification reduces as the data size increases. Index Terms - Classification, pattern recognition, hypothesis testing, decision making, machine learning, stock trading, medical diagnosis.

There is frequently interest in testing that a scalar or vector time series is I(0), possibly after first-differencing or other detrending, while the I(0) assumption is also taken for granted in autocorrelation-consistent variance estimation. We propose a test for I(0) against fractional alternatives. The test is nonparametric, and indeed makes no assumptions on spectral behaviour away from zero frequency. It seems likely to have good efficiency against fractional alternatives, relative to other nonparametric tests. The test is given large sample justification, subjected to a Monte Carlo analysis of finite sample behaviour, and applied to various empirical data series.

Anomaly detection and localization problems from noisy tomographic image projections were analyzed using a multiscale hypothesis testing technique. The class of anomalies was parameterized by intensity and position values. The number of hypotheses was directly correlated with the linear dimension of the images while the probability of false alarm was treated as a function of detection probability. Several factors hindered the analysis of large hypothesis space, including the organization of the hypothesis zooming space and data processing.

Model-selection procedures such as Akaike's information criterion, AIC, provide the basis for an attractive alternative to traditional pairwise comparisons procedures such as the Tukey HSD procedure. These proposed procedures can be set up to avoid problems, such as intransitive decisions, associated with conducting a series of correlated significance tests. Furthermore, they can be formulated for heterogeneous variance cases, can be based on non-normal distributions, and can be generalized to multivariate settings. Both univariate and bivariate examples are presented with computations conducted in a spreadsheet environment. A small-scale simulation study provides some evidence concerning the all-pairs power and patterns of decision errors for model selection using information criteria. KEY WORDS: Information criteria; Multiple comparisons; Tukey tests., 1. INTRODUCTION In order to provide the research community with an introduction to the topic of multiple-comparison procedures, Kirk (1995) described 22 techniques in the latest edition of his widely [...]

We address the problem of determining what size test set guarantees statistically significant results in a character recognition task, as a function of the expected error rate. We provide a statistical analysis showing that if, for example, the expected character error rate is around 1 percent, then, with a test set of at least 10,000 statistically independent handwritten characters (which could be obtained by taking 100 characters from each of 100 different writers), we guarantee, with 95 percent confidence, that: (1) The expected value of the character error rate is not worse than 1.25 E, where E is the empirical character error rate of the best recognizer, calculated on the test set; and (2) a difference of 0.3 E between the error rates of two recognizers is significant. We developed this framework with character recognition applications in mind, but it applies as well to speech recognition and to other pattern recognition problems. Index Terms - Pattern recognition, test set, test set size, benchmark, hypothesis testing, designed experiment, statistical significance, estimation, guaranteed estimators, recognition error.

Three parameters in quasi-periodic ultrasonic scattering predict the presence of regular structures in a given region of a tissue. The signal-to-noise ratio of phase and the chi squared statistic allow the verification of phase signal. These two parameters show peaks at demodulation frequencies. The power spectral density provides information on the level of periodicity present.

This paper derives a nonparametric procedure for testing poolability in a panel data framework. This is contrasted with alternative parametric tests which are not robust to functional misspecification. This test is illustrated with an earning equation using data from the Panel Study of Income Dynamics (PSID). Key words: Panel data; Poolability; Nonparametric; Hypothesis testing

The purpose of this paper is to examine the impact of low statistical power on the process of research replication. The traditional model of interpreting the success of replication efforts emphasizes the results of statistical significance testing. Serious errors in the interpretation of replicated studies may occur if the focus is on only statistical significance. Correct interpretation of the probability of replication must include a consideration of statistical power and related factors such as effect size and type 2 error rates. Examples are provided illustrating the importance of power, effect size, and type 2 error rates in developing a scientific consensus based on repeated experimental trials. The establishment of an empirical consensus is essential to scientific progress and may be improved by the use of aggregated effect size, where the focus is on determining the degree of replication rather than a dichotomous decision based on the results of statistical significance testing. Consensus may be delayed if investigators continue to focus on the statistical significance of individual studies and ignore the effect size and statistical power of the investigations they conduct. KEY WORDS: Experimental error; Research design; Statistical analysis., 1. INTRODUCTION In his classic text describing the development of scientific knowledge, the theoretical physicist and philosopher of science John Ziman states that the goal of science 'is a consensus [...]

Hypothesis tests are studied when the alternative hypothesis restricts the parameter space. Popular examples are multivariate one-sided tests. It is demonstrated that the likelihood ratio test is admissible and maximizes power against other choices which are arbitrarily separated from the null hypothesis. Exact findings are first derived for Gaussian linear regression models with known variance. Asymptotic analogues are developed for dynamic linear models.

The relationship between writing overdrafts and political fate among members of the 102nd Congress was studied. The data on overdraft totals and the political fate of the members were taken from the 1992 Congressional Quarterly Almanac. The benchmark value for abuse of privilege at the House bank was arbitrarily set at 100. The chi-square tests conducted for all data revealed that the greater the number of overdrafts the members make, the greater is the probability of their retiring or being defeated in the elections.

Quick restoration of power following an outage has been a challenge for power distribution engineers. Many researchers have drawn attention to this problem by proposing methods for location of outages in the distribution systems. The method presented in this paper for outage location is based on statistical hypotheses testing. The input data consists of time-of-outage measurements obtained from customer-ends. The effect of various levels of noise in the measurements on performance of the method is studied. A procedure for discarding outliers in the measurements is included. A theoretical procedure to determine the maximum noise allowed in the measurements is presented. Also, performance of the method at different locations in a test system is studied. Keywords: Distribution automation, outage location, hypothesis testing, t-statistic.