China
Africa
Explore chapters and articles related to this topic
Statistical Considerations for Ocular and Dermal Toxicology Studies
Published in David W. Hobson, Dermal and Ocular Toxicology, 2020
Robert D. Bruce, John D. Taulbee
As with grading scale data, both normal distribution methods and distribution-free methods may be useful. Although with continuous data, the usual choice would be the normal distribution methods, there are frequently instances where outlying data values or skewed distributions (a preponderance of values far from the mean in one direction) may make the adoption of normal distribution methods a bit questionable. When the experimenter has doubts about this matter, our recommendation would be either to transform the data, as illustrated below, or to use the distribution-free methods. While some statisticians complain that there is a loss in statistical efficiency entailed by using distribution-free methods when they are not needed, this loss is only about 5% (for the Wilcoxon rank sum test vs. the t test). In our experience, this slight loss in efficiency is a small price to pay since in many cases, where normality is not present, the distribution-free methods will actually be more sensitive than normal distribution methods. An exception to this would be with extremely small group sizes. For example, with two groups of size 3 or less, the smallest p value that can be achieved using the rank sum test is 0.100. In this situation, there is little point in performing a statistical test with this method; indeed, any statistical analysis may be futile in this situation.
Non Parametric Methods
Published in Marcello Pagano, Kimberlee Gauvreau, Principles of Biostatistics, 2018
Marcello Pagano, Kimberlee Gauvreau
Suppose that you have two independent populations and wish to use the rank sum test to evaluate the null hypothesis that their medians are identical. You select a sample of size 4 from the first population and a sample of size 5 from the second.
Nonparametric Comparisons of Distributions
Published in Albert Vexler, Alan D. Hutson, Xiwei Chen, Statistical Testing Strategies in the Health Sciences, 2017
Albert Vexler, Alan D. Hutson, Xiwei Chen
As a nonparametric analogue to the two-sample t-test, the Wilcoxon rank-sum test (also called the Mann–Whitney U-test or the Mann–Whitney–Wilcoxon test) can be used primarily when investigators do not want to, or cannot, assume that data distributions are known. The test itself is highly efficient and robust across a variety of parametric assumptions compared to the t-test. The key assumption for using the Wilcoxon rank-sum test is the idea of exchangeability of the observations under the null hypothesis; that is, the distribution functions for the two groups being tested are equivalent under the null hypothesis. This assumption may be generally assumed to be true in the randomized experimental setting, but may not be assumed in certain nonrandomized settings. If this assumption is not met, the Wilcoxon test might be biased.
Blood pressure patterns of hypertensive disorders of pregnancy in first and second trimester and contributing factors: a retrospective study
Published in Journal of Obstetrics and Gynaecology, 2023
Jie Ren, Zhuoran Fan, Jing Li, Yujie Wang, Junnong Zhang, Shaofang Hua
SPSS 21.0 was used for data processing. Measurement data were expressed as mean ± standard deviation and enumeration data was expressed as a percentage. For measurement data which was in accord with normal distribution and homoscedasticity, T-test was used for analysing; For heterogeneity of variance, we used an approximate T-test. For data against normal distribution, we chose the Rank sum test. The repeated measurements and 2-way ANOVA was used to observe the blood pressure variation as the gestational age increased and decided the significant difference in the same period between the normal and HDP group. For the date which was not meet the Sphericity assumption, we used Greenhouse-Geisser test. Finally, we used Chi-square test for the analysis of categorical variable data. P < 0.05 is the standard for a significant difference.
Usefulness of simultaneous impulse oscillometry and spirometry with airway response to bronchodilator in the diagnosis of asthmatic cough
Published in Journal of Asthma, 2023
Namiko Taniuchi, Mitsunori Hino, Akiko Yoshikawa, Akihiko Miyanaga, Yosuke Tanaka, Masahiro Seike, Akihiko Gemma
For continuous variables, data were expressed as mean (standard deviations) or median (interquartile range) values. We checked whether continuous variables were normally distributed. For parametric comparisons between the two groups, Student’s t-test was used. In a nonparametric test, the Wilcoxon rank-sum test was used to compare the two groups. Because of the sample size variability and the inclusion of samples that did not show normality, the Kruskal–Wallis test, a nonparametric test, was used for analysis in the three-group comparison. The Steel–Dwass test was then used to determine whether there were any significant differences between the three groups. A p values < 0.05 on both sides was considered statistically significant. The correlation between the two variables was assessed using Spearman correlation analysis. All statistical analyses were conducted using the JMP software version 11 (SAS Institute Inc., Cary, NC).
Automated prediction of COVID-19 mortality outcome using clinical and laboratory data based on hierarchical feature selection and random forest classifier
Published in Computer Methods in Biomechanics and Biomedical Engineering, 2023
Nasrin Amini, Mahdi Mahdavi, Hadi Choubdar, Atefeh Abedini, Ahmad Shalbaf, Reza Lashgari
One of the methods to evaluate the performance of binary classification algorithms is receiver operating characteristic (ROC) curve (Klawonn et al. 2011). In the ROC diagram, both sensitivity or true positive rate (TPR) and recall or false positive rate (FPR) as indicators for the performance of binary classification algorithms based on logistic regression (LR) are combined and displayed as a curve. The area under the ROC curve (AUC) is also used for the evaluation of the performance of binary classification algorithms based on given input features. AUC, as a very useful and easy-to-use framework, tells how much the model is capable of distinguishing between two classes and seeing the importance of given input features (Mamitsuka 2006). The numerical value of the AUC varies from zero to one, with numbers closer to one meaning the test method has good detection or accuracy. Finally, the Wilcoxon rank-sum test was used to evaluate the significance of the extracted features. The Wilcoxon rank-sum test is a non-parametric test for two groups whose samples are independent of each other (Fay and Proschan 2010). The probability value (p-value) of this test indicates the probability of error in accepting the validity of the observed results. Utilizing this non-parametric analysis is a common method for selecting predictive features for classification algorithms. In this study, to evaluate the performance of binary classification algorithms and select the best features for them, ROC, AUC, and p-value criteria were used.