• Journal of the Korean Data and Information Science Society
• /
• v.18 no.2
• /
• pp.525-534
• /
• 2007

It is a common practice in many research areas that multiple range tests for a between subject factor such as Tukey are applied to a within subject factor in a repeated measures design. Tukey procedure, however, sometimes detects no pairs with different means even when the hypothesis of all equal level means is rejected. This study attempts to provide a rationale for the proposition that Tukey is inappropriate post hoc procedure for a within subject factor in which the observations are correlated. We introduce two multiple range tests, Bonferroni and Scheffe, for a within subject factor and show that Bonferroni is more appropriate than Scheffe for pairwise multiple comparisons. Subsequent simulation study indicates that Tukey has significantly less power than Bonferroni in detecting actual difference between means of some pairs when the observations of a within subject factor are highly correlated.

• Journal of the Korean Society for Precision Engineering
• /
• v.18 no.12
• /
• pp.110-115
• /
• 2001

In this paper, a multiple comparison was proposed to minimize the amount of coolant. And new coolant level, called "low level", was suggested to show usefulness of multiple comparison for experiments. The amount of this levee is about 1/4 of conventional coolant amount-called "high level", and dry cutting is called "none level". Using these coolant levels, the cutting temperatures of each coolant level in turning process were analyzed by analysis of variance(ANOYA)-test and a multiple comparison. As the result of ANOVA-test, we have just known that the average of temperature of each coolant level is not equal. However by Tukey's HSD, one of multiple comparison. it was analyzed that the average of temperature of low level is similar to that of high levee and different from that of none level.erent from that of none level.

• Journal of Mechanical Science and Technology
• /
• v.18 no.12
• /
• pp.2182-2189
• /
• 2004

In the recent years, environmentally conscious design and manufacturing technologies have attracted considerable attention. The coolants, lubricants, solvents, metallic chips and discarded tools from manufacturing operations will harm our environment and the earth's ecosystem. In the present work, the Tukey method of multiple comparisons is used to select the minimum level of coolant required in a turning process. The amount of coolant is varied in 270 designed experiments and the parameters cutting temperature, surface roughness, and specific cutting energy are carefully evaluated. The effects of coolant mix ratio as well as the amount of coolant on the turning process are studied in the present work. The cutting temperature and surface roughness for different quantity of coolant are investigated by analysis of variance (ANOVA) - test and a multiple comparison method. ANOVA-test results signify that the average tool temperature and surface roughness depend on the amount of coolant. Based on Tukey's Honestly Significant Difference (HSD) method, one of the multiple comparison methods, the minimum level of coolant is 1.0 L/min with 2% mix ratio in the aspect of controlling tool temperature. F-test concludes that the amount of coolant used does not have any significant effect on specific cutting energy. Finally, Tukey method ascertains that 0.5 L/min with 6% mix ratio is the minimum level of coolant required in turning process without any serious degradation of the surface finish. Considering all aspects of cutting, the minimum coolant required is 1.0 L/min with 6% mix ratio. It is merely half the coolant currently used i.e. 2.0 L/min with 10% mix ratio. Minimal use of coolant not only economically desirable for reducing manufacturing cost but also it imparts fewer hazards to human health. Also, sparing use of coolant will eventually transform the turning process into a more environment-conscious manufacturing process.

• Communications for Statistical Applications and Methods
• /
• v.16 no.1
• /
• pp.75-83
• /
• 2009

Parallelism tests are proposed for slope in the simple linear regression models. In this paper, we suggest the parametric test using HSD testing method (Tukey,1953) and distribution-free test using Kruskal-wallis (1952) for more than three slopes. Monte Carlo simulation study is adapted to compare the power of the proposed methods with Wilks' Lambda multivariate procedure.

• Journal of Korean Society for Quality Management
• /
• v.45 no.4
• /
• pp.933-942
• /
• 2017

Purpose: The purpose of this paper is to evaluate the performance quality of the open source DBMSs. Performance quality is defined as processing time for Join queries. Query processing time is measured and compared in the most widely used open source DBMSs and commercial DBMS. Methods: By varying the number of tuples of two relations to be joined, the average processing time(seconds) of a Join query in each DBMS was obtained experimentally. ANOVA and Tukey HSD test were used in order to compare the performance quality of DBMSs. Results: There was a significant difference between the performance qualities of the three DBMSs at all experimental levels where the number of tuples was 100, 1,000, 2,000, 10,000, and 50,000. As a result of the Tukey HSD test, two open source DBMSs (MariaDB, MySQL) were classified in the same group only at the tuple level of 100. The commercial DBMS (MS-SQL Server) belonged to another group. At level of more than 1,000 tuples, all three DBMSs belonged to different groups. Conclusion: Within the open source DBMS group, MariaDB showed the better performance quality except for a small number of tuples. Thus the results show that MariaDB can be the alternative to MySQL which is currently most widely used. Between open source DBMS and commercial DBMS groups, MS-SQL Server always shows the best performance quality, but the less number of tuples, the less the difference.

• Journal of Technologic Dentistry
• /
• v.18 no.1
• /
• pp.67-72
• /
• 1996

This test is verifying for influence of W/P ratio of investment upon dental prosthesis Firstly we made 40 MOD die and divide 4 groups, each group take 0.2 different w/p ratio grade such as 0.35, 0.37, 0.39 and 0.41. The method of data analysis applied were by ANOVA, Tukey test and Mann-Whiney and Kruskal Wallis test. The finding of this study were as follows : 1. The 0.37 w/p ratio group showed most accurte margin due to $15.8{\mu}m$ gap between margin and die which keep up with manufacture's instruction and 0.35 w/p ratio group was followed as second accuracy group o 새 $48.3{\mu}m$ gap between margin and die, 0.39 w/p ratio group showed $101{\mu}m$ gap and 0.41 w/p ratio group showed $129.8{\mu}m$ gap. 2. As to the relationship of the margin accuracy between 4 different grade of w/p ratio groups, each group was statistically significant(P<0.01). Also the all groups were statistically significant except between 0.35 and 0.37 groups between 0.39 and 0.41 groups, and between 0.35 and 0.39 groups. 3. A consequence of Mann-Whitney & Kruskal-Wallis test for marginal accuracy between 4 different grade of w/p ratio was statistically significant as same as above ANOVA test result. Also Tukey test for verifying similarity from each group showed same as above.

• International Journal of Highway Engineering
• /
• v.15 no.6
• /
• pp.61-68
• /
• 2013

PURPOSES : The purpose of this study is to study the noise reducing effect of Micro Surfacing by comparing general asphalt, longitudinal tining and Slurry Seal. METHODS : This study measures vehicles' noise of each section by the NCPX method that can measure noise between the road surface and the tire at the field. Total sound pressure and sound pressure level by the 1/3 octave band frequency are calculated through the field data of each section. Total sound pressure level is compared by ANOVA test statistically. After ANOVA test, post-hoc test is conducted to know mean difference of surface type by Tukey. RESULTS : As the result of CPB analysis to confirm sound pressure levels by frequency, it was shown that sound pressure levels by frequency are totally similar except for those of frequency bands between 100Hz and 500Hz. The result of ANOVA test and post-hoc test, it was shown that sections of surface type have a difference. The result of Micro Surfacing was lower 2~5dB(A) than other surface type. CONCLUSIONS : It is found that the noise reduction performance of Micro Surfacing was better than other surface type.

• The Journal of Korean Academy of Prosthodontics
• /
• v.54 no.3
• /
• pp.218-225
• /
• 2016

Purpose: The purpose of this study was to evaluate the push-out bond strength of glass-fiber post cemented with different adhesive systems and surface treatments. Materials and methods: 160 tooth samples made from 48 human maxillary single-rooted teeth with similar root length were divided into 4 groups according to the adhesive system (no adhesive, Adper Single Bond 2, Clearfil SE Bond, Clearfil S3). Each group had 4 subgroups according to the post surface treatment methods (no treatment, sandblast, silane, sandblast and silane). Posts (Parapost Fiber White) were cemented with Rely X Unicem. The teeth were sectioned perpendicular to their long axis into 1-mm thick sections. The push-out tests was performed at a speed of 0.5 mm/min. The results were evaluated by 2-way ANOVA, 1-way ANOVA and multiple comparison procedures (Tukey test) (${\alpha}=0.05$). Results: Tukey test showed that the adhesive system significantly influenced the push-out strength. The Clearfil SE Bond group showed the highest value. Post surface treatments showed no significant effect. Conclusion: Bond strength of glass-fiber post cemented with self-adhesive resin cement using Clearfil SE Bond showed significantly higher values compared to other adhesive systems.

• Korean Journal of Ecology and Environment
• /
• v.39 no.2 s.116
• /
• pp.209-218
• /
• 2006

This study was to evaluate microcystin production by two strains of cyanobacteria (Microcystis aeruginosa and Planktothrix agardhii) in response to three different levels of direct (0,4,8 inds.) or indirect (0,25, 50% of zooplankton culture media filtrate) exposures to zooplankton (Daphnia magna and Moina macrocopa). The cell biomass and intracellular microcystin (MC) were measured everyday. The survival rates of zooplankton were evaluated for daily intervals for the direct exposure. The intracellular MC produced peaked on the day 3 or 4, and then decreased over the both exposure experiment. In the direct experiment, the MC values were significantly different among the control and zooplankton treatments (ZT; repeated measures-ANOVA: P< 0.039). The MC contents of P. agardhii strain (No.204) were significantly higher (Tukey test, P< 0.082) in ZT2 (8 inds.) than in ZT2 (4 inds.). On the peak day, the intracellular MC exposed to both zooplanktons was significantly higher than the control (One-way ANOVA, P< 0.021). Higher zooplankton survivals were observed in the M. aeruginosa strain (No. 111) rather than in high toxic P. agardhii strain. In the indirect experiment, the intracellular MC of the M. aeruginosa strain was significantly different among the control and zooplankton culture media filtrate (ZCMF)treatments (rm-ANOVA: P<0.004), The MC exposed ZCMF2 (50%) were significantly higher than in ZCMFI (25%: Tukey test, P< 0.025) for both strains. This study strongly supports the induced-defensive MC production of potentially toxic cyanobacteria in response to the presence of zooplankton.

• Communications for Statistical Applications and Methods
• /
• v.30 no.2
• /
• pp.163-178
• /
• 2023

GARCH-X(1, 1) model specifies that conditional variance follows an AR(1) process and includes a past exogenous variable. This study proposes a new class from that model by allowing a more general (non-linear) variance function to follow an AR(1) process. The functions applied to the variance equation include exponential, Tukey's ladder, and Yeo-Johnson transformations. In the framework of normal and student-t distributions for return errors, the empirical analysis focuses on two stock indices data in developed countries (FTSE100 and SP500) over the daily period from January 2000 to December 2020. This study uses 10-minute realized volatility as the exogenous component. The parameters of considered models are estimated using the adaptive random walk metropolis method in the Monte Carlo Markov chain algorithm and implemented in the Matlab program. The 95% highest posterior density intervals show that the three transformations are significant for the GARCHX(1, 1) model. In general, based on the Akaike information criterion, the GARCH-X(1, 1) model that has return errors with student-t distribution and variance transformed by Tukey's ladder function provides the best data fit. In forecasting value-at-risk with the 95% confidence level, the Christoffersen's independence test suggest that non-linear models is the most suitable for modeling return data, especially model with the Tukey's ladder transformation.