• Journal of the Korean Data and Information Science Society
• /
• v.14 no.3
• /
• pp.673-685
• /
• 2003

In this paper we consider the problem of identifying and testing outliers in linear regression. First we consider the problem for testing the null hypothesis of no outliers. A test based on the ratio of two residual scale estimates is proposed. We show the asymptotic distribution of the test statistics by Monte Carlo simulation and investigate its properties. Next we consider the problem of identifying the outliers. A forward sequential procedure using the suggested test is proposed and shown to perform fairly well. Unlike other forward procedures, the present one is unaffected by masking and swamping effects because the test statistic is based on robust scale estimate.

• Journal of the Korean Data and Information Science Society
• /
• v.16 no.4
• /
• pp.1053-1066
• /
• 2005

In this paper we consider the problem of identifying and testing outliers in linear regression. First we consider the use of the so-called scale ratio tests for testing the null hypothesis of no outliers. This test is based on the ratio of two residual scale estimates. We show the asymptotic distribution of the test statistics and investigate its properties. Next we consider the problem of identifying the outliers. A forward sequential procedure using the suggested test is proposed. The new method is compared with classical procedure in the real data example. Unlike other forward procedures, the present one is unaffected by masking and swamping effects because the test statistic is based on robust scale estimate.

• Communications for Statistical Applications and Methods
• /
• v.2 no.1
• /
• pp.122-136
• /
• 1995

The objective of this research is to investigate the problem of goodness-of-fit testing based on nonparametric density estimation with a data-driven smoothing parameter. The small and large sample properties of a new test statistic $\hat{\lambda_a}$ is investigated. The test statistic $\hat{\lambda_a}$ is itself a smoothing parameter which is selected to minimize an estimated MISE for a truncated series estimator of the comparison density function. Therefore, this test statistic leads immediately to a point estimate of the density function th the event that $H_0$ is rejected. The limiting distribution of $\hat{\lambda_a}$ is obtained under the null hypothesis. It is also shown that this test is consistent against fixed alternatives.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.6
• /
• pp.809-813
• /
• 2009

We want to show that the construct of best fuzzy tests for certain fuzzy situations of Poisson distribution. Due to Neyman and Pearson theorem, if we have ${\theta}_0$ and ${\theta}_1$ be distinct fuzzy values of ${\Omega}=\{{\theta}\;:\;{\theta}\;=\;{\theta}_0,\;{\theta}_1\}$ such that $L({\theta}_0\;:\;X)/L({\theta}_1\;:\;X)$ < k, then k is a fuzzy number. For each fuzzy random samples point $X\;{\subset}\;C$, we have most power test for fuzzy critical region C by agreement index.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.6
• /
• pp.592-596
• /
• 2014

We shall consider fuzzy hypotheses test for signed-rank Wilcoxon fuzzy test by fuzzy difference on rate of internal division. Fundamental to these discussion are fuzzy number data and Wilcoxon signed-rank fuzzy test of a fuzzy hypothesis $H_{f0}$ which is based upon a fuzzy statistics whose distribution does not depend upon the specified distribution or any parameters.

• Research in Community and Public Health Nursing
• /
• v.7 no.2
• /
• pp.333-348
• /
• 1996

The purpose of this study was to examine the effect of nursing information on the level of need fulfillment and anxiety of the emergency patient's family members. The quasi-experimental study was designed using a noneqivalent control group non-synchronized design. During the first period, 30 subjects were assigned to the control group and 25 to the experimental group at a late period. The experimental group was provided with nursing information via guide booklet designed by the researcher. The control group received only routine care. Data was collected from January 31 to April 16 in 1996 at the K hospital in Taegu and analysed by chi-square test, t-test, ANOVA and Pearson correlation with SAS program. The instruments used for this study were the Family Needs Scale developed by Jung and the State-Trait Anxiety Inventory developed by Spielberger. The results of this study were summarized as follows : 1. The first hypothesis that the family members who received nursing information will have greater need fulfillment than family members who did not receive nursing information was supported. 2. The second hypothesis that the family members who received nursing information will have lower anxiety level than family members who did not receive nursing information was not supported. 3. The third hypothesis that the more the need of family member of emergency patient was met, the lower the anxiety level, was not supported. In conclusion, it was proved that nursing information about the emergency room provides family members with more need fulfillment, but did not decrease the anxiety level.

• Cartoon and Animation Studies
• /
• s.40
• /
• pp.337-365
• /
• 2015

The objective of this study proposes the new user's addiction model in 'Social Network Games' (SNGs). Research model is derived from the separation of two characteristics. First one is logical characteristics that includes 'Functional' (F), 'Keystroke' (K), and 'Goal' (G). Second one is feeling characteristics that consists a few factors such as 'Emotion' (E), 'Social' (S), and 'Affection' (A). For the pre-test, a total of 30 participants responded to survey in order to inspect the fitness of research questionnaire, roughly validity of the proposed model, and the direction of this reseach. After that for the main test, a total 300 users participated in this research. The final number of effective participants were 261 because 39 were insincere respondents and without playing SNGs who were excluded. Then we examined the measurement model by performing 'Partial Least Squares - Structural Equation Modeling' (PLS-SEM) analysis to test the research hypothesis empirically. The results of the measurement and structural model test lend support to the proposed research model by providing a good fit to the construct data. Interestingly, the model showed the significant effects of the interaction between eleven hypothesis(H1,H2,H3,H4,H5,H6,H7,H8,H9,H10, H12). Only one hypothesis decision t-value not supported that is involved the relationship between SNGs Addiction and Keystroke, H11(1.193). This research expect to contributes to an exploratory SNGs research to clarify the base of addition and will aids understanding of users' behavior associated with SNGs development.

• The Journal of the Korea Contents Association
• /
• v.14 no.6
• /
• pp.367-384
• /
• 2014

This study addresses an empirical issue which has been received little attention in the contemporary finance literature: To identify any financial determinants of the profitability indices for the firms belonging to the Korean chaebol. Three hypotheses of concern were postulated and tested for the sample firms covering the periods of the pre-and post-financial global crises. Regarding the results on the 1st hypothesis test of characterizing any financial profiles for the firms (belonging to the chaebols) by estimating a legitimate panel data model: the present study found the statistically significant relationships of the explanatory variables (BVLEVl, MVLEVl, MV/BV, RISK, FCFF and FOS) with the book-value based profitability ratio: while the market-valued profitability index was explained only by BVLEV2. Regarding the 2nd hypothesis test for the profitability of the sample firms at the industry level: the chaebol firms in the chemical and the food industries overall positioned themselves into the top ranks in order, which was tested by the ANCOVA and the Tukey multiple comparison procedure. Finally: on the 3rd hypothesis test for the 'adjusted' Dupont system, only two such as the 'operating margin' and the 'asset turnover' showed their significant effects between the chaebol firms and their counterparts in both the (parametric) independent samples t-test and the (nonparametric) Wilcoxon-Mann-Whitney statistics.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.8
• /
• pp.87-97
• /
• 2020

The aim of this study was to examine empirically the causal relationship between significant adverse outcomes of elderly frailty. The 2017 National Survey on Old Adults was used to test this, but only nine variables necessary for the hypothesis test were abstracted from it. The results of the hypothesis test are summarized as follows. The sub-variables of adverse outcomes, such as disabilities in daily living activities, disabilities in instrumental daily living activities, and health care utilization, were analyzed empirically by inputting sub-variables of six frailty factors into hierarchical regression analysis. In the final model, the physical activities, emotional control, and social support were verified as significant variables influencing the disabilities in daily living activities. These included the following: physical activities, sensory functions, cognition, emotional control and social support as those influencing disabilities in the instrumental daily living activities; and nutrition, physical activities, sensory functions, emotional control, and social support as those influencing the health care utilization. Based on the results of the hypothesis test, it is necessary to develop and implement personalized health management programs that specifically identify elderly weakness factors that affect adverse outcomes, such as disabilities or health care utilization.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.233-236
• /
• 2014

The purpose of this study is for the students who are naturally diverse in personality and behavior in the alternative school to examine the effect of the computer subject achievement and how to increase the student's interest in learning by means of the implementation of Howard Gardener, the Multiple-Intelligence Hypothesis. To verify this study, two groups of sophomore students from an Alternative High School located in Hwasung : The Experimental group consisted of 20 students who would participate in the Computer Science program by learning or applying new skills, based on the Multiple Intelligence Hypothesis. The Comparison Group consisted of 19 students who would follow the standardized Computer Science teaching and lesson plan. During five weeks, the multiple-Intelligence effect was executed in the search groups comparing the differences of the pre-test and pro-test. Pro-test showed following results: Firstly, the experiment search group showed a higher significant improvement in the subject achievement, compared to the comparison group. Secondly, the experiment search group showed a much higher increase in the interest of learning, compared to the comparison group. Thirdly, the experiment search group did not show any differences of the social development compared to the comparison group.