• The Korean Journal of Applied Statistics
• /
• v.20 no.1
• /
• pp.79-89
• /
• 2007

A test for normality introduced by Arizono and Ohta(1989) is based on fullback-Leibler discrimination information. The test statistic is derived from the discrimination information estimated using sample entropy of Vasicek(1976) and the maximum likelihood estimator of the variance. However, these estimators are biased and so it is reasonable to make use of unbiased estimators to accurately estimate the discrimination information. In this paper, Arizono-Ohta test for normality is improved. The derived test statistic is based on the bias-corrected entropy estimator and the uniformly minimum variance unbiased estimator of the variance. The properties of the improved KL test are investigated and Monte Carlo simulation is performed for power comparison.

• The Journal of Asian Finance, Economics and Business
• /
• v.7 no.10
• /
• pp.523-530
• /
• 2020

The efficient market hypothesis explains the random walk hypothesis suggesting that stock prices are independent of each other, hence, it is impossible to earn abnormal profits. The positive effect of a well-functioning and highly efficient stock market on the performance of an economy motivated the Philippine Stock Exchange to pursue massive modernization initiatives. This research provides evidence of the existence of random walk in the Philippine stock market employing the Augmented Dickey-Fuller (1981) and Phillips-Perron (1988) unit root tests, the Lo-MacKinlay's (1988) conventional variance ratio test, and Chow-Denning's (1993) simple multiple variance ratio test. Results of the ADF and PP unit root tests confirm the necessary condition for a random walk. The Chow-Denning (1993) maximum /z/ statistic and the Wald test statistic as in Richardson and Smith (1991) for the joint hypotheses and the Lo and MacKinlay (1988) individual statistics variance ratio test generally accepted the null hypothesis of a random walk. That is, the unit root and variance ratio tests consistently indicate that the null hypothesis of random walk cannot be rejected. The existence of a random walk in weak-form efficiency can be attributed to market liquidity as a result of continuous development and modernization of the Philippine equity market.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.1
• /
• pp.73-83
• /
• 2013

Among class quantitative tests for the credit rating systems, the credit rating tests for calibration are to test the class homogeneous differences between observed and predicted probabilities. For one time period, binomial test and chi-square test are included, and normal test and extended traffic lights test are also contained for several time peroids. In this work, we consider real data in which there exists correlation among variables, so that these test methods could be applied to the credit rating systems as well as various kinds of the class data such as BWT data and FSI data.

• Journal of the Korean Statistical Society
• /
• v.30 no.3
• /
• pp.499-509
• /
• 2001

We investigate the influence of observations on a test of additional information about discrimination using the influence function and the derivative influence measures. the influence function for the test statistic is derived and this sample versions are used for influence analysis. The derivative influence measures for the test statistic under a perturbation scheme are derived. It will be seen that the influence function method and the derivative influence measures yield the same result. Furthermore, we will derive the relationships between the influence function and the derivative influence measures when the sample size is large. an illustrative example is given and we will compare the results provided by the influence function method and the derivative influence measures.

• Journal of the Korea Safety Management & Science
• /
• v.10 no.2
• /
• pp.205-210
• /
• 2008

The research interprets the principles of sampling error design for quality statistics models such as hypothesis test, interval estimation, control charts and acceptance sampling. Introducing the proper discussions of the design of significance level according to the use of hypothesis test, then it presents two methods to interpret significance by Neyman-Pearson and Fisher. Second point of the study proposes the design of confidence level for interval estimation by Bayesian confidence set, frequentist confidential set and fiducial interval. Third, the content also indicates the design of type I error and type II error considering both productivity and customer claim for control chart. Finally, the study reflects the design of producer's risk with operating charistictics curve, screening and switch rules for the purpose of purchasing and subcontraction.

• Communications for Statistical Applications and Methods
• /
• v.10 no.2
• /
• pp.277-289
• /
• 2003

We propose modified exact inferential methods in logistic regression model. Exact conditional distribution in logistic regression model is often highly discrete, and ordinary exact inference in logistic regression is conservative, because of the discreteness of the distribution. For the exact inference in logistic regression model we utilize the modified P-value. The modified P-value can not exceed the ordinary P-value, so the test of size $\alpha$ based on the modified P-value is less conservative. The modified exact confidence interval maintains at least a fixed confidence level but tends to be much narrower. The approach inverts results of a test with a modified P-value utilizing the test statistic and table probabilities in logistic regression model.

• Journal of Korean Society for Quality Management
• /
• v.21 no.2
• /
• pp.162-169
• /
• 1993

For the problem of testing the parallelism of several regression lines against ordered alternatives, two test statistics and proposed and examined. The proposed statistics are linear combinations of robust estimators of slope parameters, which are modifications of the Adichie (1976) test based on scores. The asymptotic null variances of the proposed states tics are estimated by the kernel density estimation methods. The proposed tests are compared with the Adichie's test in terms of asymptotic relative efficiency and small-sample powers.

• Communications for Statistical Applications and Methods
• /
• v.3 no.3
• /
• pp.205-214
• /
• 1996

As flexible Bayesian test criteria for nested point null hypotheses of multiple regression coefficients, partial and overall Bayes factors are introduced under a class of intuitively meaningful prior. The criteria lead to a simple method for considering different prior beliefs on the subspaces that constitute a partition of the coefficient parameter space. A couple of tests are suggested based on the criteria. It is shown that they enable us to obtain pairwise comparisons of hypotheses of the partitioned subspaces. Through a Monte Carlo simulation, performance of the tests based on the criteria are compared with the usual Bayesian test (based on Bayes factor)in terms of their respective powers.

• The Korean Journal of Applied Statistics
• /
• v.28 no.3
• /
• pp.573-581
• /
• 2015

We study problems of sample size determination for one-sample location tests. A simulation study shows that sample size calculations based on approximated distribution do not achieve the nominal level of power. We investigate sample size determinations based on exact distribution and with a power that attains the nominal level.

• The Korean Journal of Applied Statistics
• /
• v.22 no.1
• /
• pp.205-220
• /
• 2009

For modeling skewed semicircular data, we derive new family of the exponential distributions. We extend it to the l-axial exponential distribution by a transformation for modeling any arc of arbitrary length. It is straightforward to generate samples from the f-axial exponential distribution. Asymptotic result reveals two things. The first is that linear exponential distribution can be used to approximate the l-axial exponential distribution. The second is that the l-axial exponential distribution has the asymptotic memoryless property though it doesn't have strict memoryless property. Some trigonometric moments are also derived in closed forms. Maximum likelihood estimation is adopted to estimate model parameters. Some hypotheses tests and confidence intervals are also developed. The Kolmogorov-Smirnov test is adopted for goodness of fit test of the l-axial exponential distribution. We finally obtain a bivariate version of two kinds of the l-axial exponential distributions.