DOCSLIB.ORG

Hypothesis Testing with the Bootstrap

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Hypothesis Testing with the Bootstrap Hypothesis Testing with the Bootstrap Noa Haas Statistics M.Sc. Seminar, Spring 2017 Bootstrap and Resampling Methods Bootstrap Hypothesis Testing A bootstrap hypothesis test starts with a test statistic - 푡(풙) (not necessary an estimate of a parameter). We seek an achieved significance level ∗ 퐴푆퐿 = 푃푟표푏퐻0 푡 풙 ≥ 푡(풙) Where the random variable 풙∗ has a distribution specified by the null hypothesis 퐻0 - denote as 퐹0. Bootstrap hypothesis testing uses a “plug-in” style to estimate 퐹0. The Two-Sample Problem We observe two independent random samples: 퐹 → 풛 = 푧1, 푧2, … , 푧푛 푛푑푒푝푒푛푑푒푛푡푙푦 표푓 퐺 → 풚 = 푦1, 푦2, … , 푦푚 And we wish to test the null hypothesis of no difference between F and G, 퐻0: 퐹 = 퐺 Bootstrap Hypothesis Testing 퐹 = 퐺 • Denote the combined sample by 풙, and its empirical distribution by 퐹 0. • Under 퐻0, 퐹 0 provides a non parametric estimate for the common population that gave rise to both 풛 and 풚. 1. Draw 푩 samples of size 푛 + 푚 with replacement from 풙. Call the first n observations 풛∗ and the remaining 푚 – 풚∗ 2. Evaluate 푡(∙) on each sample - 푡(풙∗풃) 3. Approximate 퐴푆퐿푏표표푡 by ∗풃 퐴푆퐿 푏표표푡 = # 푡 풙 ≥ 푡 풙 /퐵 ∗풃 * In the case that large values of 푡(풙 ) are evidence against 퐻0 Bootstrap Hypothesis Testing 퐹 = 퐺 on the Mouse Data A histogram of bootstrap replications of 푡 풙 = 푧 − 푦 for testing 퐻0: 퐹 = 퐺 on the mouse data. The proportion of values greater than 30.63 is .121. Calculating 푧 − 푦 푡 풙 = 휎 1 푛 + 1 푚 (approximate pivotal) for the same replications produced 퐴푆퐿 푏표표푡 = .128 Testing Equality of Means • Instead of testing 퐻0: 퐹 = 퐺, we wish to test H0: 휇푧 = 휇푦, without assuming equal variances. We need estimates of 퐹 and 퐺 that use only the assumption of common mean 1. Define points 푧 = 푧 − 푧 + 푥 , = 1, … , 푛, and 푦 = 푦 − 푦 + 푥 , = 1, … , 푚. The empirical distributions of 풛 and 풚 shares a common mean. 2. Draw 푩 bootstrap samples with replacement 풛∗, 풚∗ from 푧1 , 푧2 , … , 푧푛 and 푦 1, 푦 2, … , 푦 푚 respectivly 3. Evaluate 푡(∙) on each sample - ∗ ∗ ∗풃 푧 − 푦 푡 풙 = ∗ ∗ 휎 푧 1 푛 + 휎 푦 1 푚 4. Approximate 퐴푆퐿푏표표푡 by ∗풃 퐴푆퐿 푏표표푡 = # 푡 풙 ≥ 푡 풙 /퐵 Permutation Test VS Bootstrap Hypothesis Testing • Accuracy: In the two-sample problem, 퐴푆퐿푝푒푟푚 is the exact probability of obtaining a test statistic as extreme as the one observed. In contrast, the bootstrap explicitly samples from estimated probability mechanism. 퐴푆퐿 푏표표푡 has no interpretation as an exact probability. • Flexibility: When special symmetry isn’t required, the bootstrap testing can be applied much more generally than the permutation test. (Like in the two sample problem – permutation test is limited to 퐻0: 퐹 = 퐺, or in the one-sample problem) The One-Sample Problem We observe a random sample: 퐹 → 풛 = 푧1, 푧2, … , 푧푛 And we wish to test whether the mean of the population equals to some predetermine value 휇0 − 퐻0: 휇푧 = 휇0 Bootstrap Hypothesis Testing 휇푧 = 휇0 What is the appropriate way to estimate the null distribution? The empirical distribution 퐹 is not an appropriate estimation, because it does not obey 퐻0. As before, we can use the empirical distribution of the points: 푧 = 푧 − 푧 + 휇0, = 1, … , 푛 Which has a mean of 휇0. Bootstrap Hypothesis Testing 휇푧 = 휇0 The test will be based on the approximate 푧 −휇 distribution of the test statistic 푡 풛 = 0 휎 / 푛 ∗ ∗ We sample 푩 times 푧1 , … , 푧푛 with replacement from 푧1 , … , 푧푛 , and for each sample compute 푧 − 휇0 푡 풛 ∗ = 휎 / 푛 And the estimated ASL is given by ∗풃 퐴푆퐿 푏표표푡 = # 푡 풛 ≥ 푡 풛 /퐵 ∗풃 * In the case that large values of 푡 풛 are evidence against 퐻0 Testing 휇푧 = 휇0 on the Mouse Data Taking 휇0 = 129, the observed value of the test statistic is 86.9 − 129 푡 풛 = = −1.67 66.8/ 7 (When estimating 휎 with the unbiased estimator for standard deviation). For 94 of 1000 bootstrap samples, 푡 풛 ∗ was smaller than -1.67, and therefor 퐴푆퐿 푏표표푡 = .094 For reference, the student’s t-test result for the same null hypothesis on that data gives us 42.1 퐴푆퐿 = 푃푟표푏 푡6 < − = 0.07 66.8/ 7 Testing Multimodality of a Population A mode is defined to be a local maximum or “bump” of the population density The data: 푥1, … , 푥485 Mexican stamps’ thickness from 1872. The number of modes is suggestive of the number of distinct type of paper used in the printing. Testing Multimodality of a Population Since the histogram is not smooth, it is difficult to tell from it whether there are more than one mode. A Gaussian kernel density with window size ℎ estimate can be used in order to obtain a smoother estimate: 푛 1 푡 − 푥 푓 푡; ℎ = 휙 푛ℎ ℎ =1 As 풉 increases, the number of modes in the density estimate is non-increasing Testing Multimodality of a Population The null hypothesis: 퐻0: 푛푢푚푏푒푟 표푓 푚표푑푒푠 = 1 Versus 푛푢푚푏푒푟 표푓 푚표푑푒푠 > 1. Since the number of modes decreases as ℎ increases, there is a smallest value of ℎ such that 푓 푡; ℎ has one mode. Call it ℎ 1. In our case, ℎ 1 ≈ .0068. Testing Multimodality of a Population It seems reasonable to use 푓(푡; ℎ1) as the estimated null distribution for our test of 퐻0. It is the density estimate that uses least amount of smoothing among all estimated with one mode (conservative). A small adjustment to 푓 is needed because the formula artificially 2 increases the variance of the estimate with ℎ1. Let 푔 ∙; ℎ1 be the rescale estimate, that imposes variance equal to the sample variance. A natural choice for a test statistic is ℎ1 - a large value of ℎ1 is evidence against 퐻0. Putting all of this together, the achieved significance level is ∗ 퐴푆퐿 = 푃푟표푏 ℎ > ℎ 푏표표푡 ∙;ℎ 1 1 1 ∗ Where each bootstrap sample 풙 is drawn from 푔 ∙; ℎ1 Testing Multimodality of a Population The sampling from 푔 ∙; ℎ1 is given by: 1 − ∗ 2 2 2 ∗ 푥 = 푥 + 1 + ℎ1 휎 푦 − 푥 + ℎ1휖 ; = 1, … , 푛 ∗ ∗ Where 푦1, … , 푦푛 are sampled with replacement from 푥1, … , 푥푛, and 휖 are standard normal random variables. (called smoothed bootstrap) In the stamps data, out of 500 bootstrap samples, none had ∗ ℎ1 > .0068, so 퐴푆퐿 푏표표푡 = 0. The results can be interpreted in sequential manner, moving on to higher values of the least amount of modes. (Silverman 1981) When testing the same for 퐻0: 푛푢푚푏푒푟 표푓 푚표푑푒푠 = 2, 146 ∗ samples out of 500 had ℎ2 > .0033, which translates to 퐴푆퐿 푏표표푡 = 0.292. In our case, the inference process will end here. Summary A bootstrap hypothesis test is carried out using the followings: a) A test statistic 푡(풙) b) An approximate null distribution 퐹 0 for the data under 퐻0 ∗ Given these, we generate 퐵 bootstrap values of 푡(풙 ) under 퐹 0 and estimate the achieved significance level by ∗풃 퐴푆퐿 푏표표푡 = # 푡 풙 ≥ 푡(풙) /퐵 The choice of test statistic 푡 풙 and the estimate of the null distribution will determine the power of the test. In the stamp example, if the actual population density is bimodal, but the Gaussian kernel density does not approximate it accurately, then the suggested test will not have high power. Bootstrap tests are useful when the alternative hypothesis is not well specified. In cases where there is parametric alternative hypothesis, likelihood or Bayesian methods might be preferable. .
Load more

Recommended publications
  • Analysis of Imbalance Strategies Recommendation Using a Meta-Learning Approach
    7th ICML Workshop on Automated Machine Learning (2020) Analysis of Imbalance Strategies Recommendation using a Meta-Learning Approach Afonso Jos´eCosta [email protected] CISUC, Department of Informatics Engineering, University of Coimbra, Portugal Miriam Seoane Santos [email protected] CISUC, Department of Informatics Engineering, University of Coimbra, Portugal Carlos Soares [email protected] Fraunhofer AICOS and LIACC, Faculty of Engineering, University of Porto, Portugal Pedro Henriques Abreu [email protected] CISUC, Department of Informatics Engineering, University of Coimbra, Portugal Abstract Class imbalance is known to degrade the performance of classifiers. To address this is- sue, imbalance strategies, such as oversampling or undersampling, are often employed to balance the class distribution of datasets. However, although several strategies have been proposed throughout the years, there is no one-fits-all solution for imbalanced datasets. In this context, meta-learning arises a way to recommend proper strategies to handle class imbalance, based on data characteristics. Nonetheless, in related work, recommendation- based systems do not focus on how the recommendation process is conducted, thus lacking interpretable knowledge. In this paper, several meta-characteristics are identified in order to provide interpretable knowledge to guide the selection of imbalance strategies, using Ex- ceptional Preferences Mining. The experimental setup considers 163 real-world imbalanced datasets, preprocessed with 9 well-known resampling algorithms. Our findings elaborate on the identification of meta-characteristics that demonstrate when using preprocessing algo- rithms is advantageous for the problem and when maintaining the original dataset benefits classification performance. 1. Introduction Class imbalance is known to severely affect the data quality.
    [Show full text]
  • Particle Filtering-Based Maximum Likelihood Estimation for Financial Parameter Estimation
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Spiral - Imperial College Digital Repository Particle Filtering-based Maximum Likelihood Estimation for Financial Parameter Estimation Jinzhe Yang Binghuan Lin Wayne Luk Terence Nahar Imperial College & SWIP Techila Technologies Ltd Imperial College SWIP London & Edinburgh, UK Tampere University of Technology London, UK London, UK [email protected] Tampere, Finland [email protected] [email protected] [email protected] Abstract—This paper presents a novel method for estimating result, which cannot meet runtime requirement in practice. In parameters of financial models with jump diffusions. It is a addition, high performance computing (HPC) technologies are Particle Filter based Maximum Likelihood Estimation process, still new to the finance industry. We provide a PF based MLE which uses particle streams to enable efficient evaluation of con- straints and weights. We also provide a CPU-FPGA collaborative solution to estimate parameters of jump diffusion models, and design for parameter estimation of Stochastic Volatility with we utilise HPC technology to speedup the estimation scheme Correlated and Contemporaneous Jumps model as a case study. so that it would be acceptable for practical use. The result is evaluated by comparing with a CPU and a cloud This paper presents the algorithm development, as well computing platform. We show 14 times speed up for the FPGA as the pipeline design solution in FPGA technology, of PF design compared with the CPU, and similar speedup but better convergence compared with an alternative parallelisation scheme based MLE for parameter estimation of Stochastic Volatility using Techila Middleware on a multi-CPU environment.
    [Show full text]
  • Resampling Hypothesis Tests for Autocorrelated Fields
    JANUARY 1997 WILKS 65 Resampling Hypothesis Tests for Autocorrelated Fields D. S. WILKS Department of Soil, Crop and Atmospheric Sciences, Cornell University, Ithaca, New York (Manuscript received 5 March 1996, in ®nal form 10 May 1996) ABSTRACT Presently employed hypothesis tests for multivariate geophysical data (e.g., climatic ®elds) require the as- sumption that either the data are serially uncorrelated, or spatially uncorrelated, or both. Good methods have been developed to deal with temporal correlation, but generalization of these methods to multivariate problems involving spatial correlation has been problematic, particularly when (as is often the case) sample sizes are small relative to the dimension of the data vectors. Spatial correlation has been handled successfully by resampling methods when the temporal correlation can be neglected, at least according to the null hypothesis. This paper describes the construction of resampling tests for differences of means that account simultaneously for temporal and spatial correlation. First, univariate tests are derived that respect temporal correlation in the data, using the relatively new concept of ``moving blocks'' bootstrap resampling. These tests perform accurately for small samples and are nearly as powerful as existing alternatives. Simultaneous application of these univariate resam- pling tests to elements of data vectors (or ®elds) yields a powerful (i.e., sensitive) multivariate test in which the cross correlation between elements of the data vectors is successfully captured by the resampling, rather than through explicit modeling and estimation. 1. Introduction dle portion of Fig. 1. First, standard tests, including the t test, rest on the assumption that the underlying data It is often of interest to compare two datasets for are composed of independent samples from their parent differences with respect to particular attributes.
    [Show full text]
  • Sampling and Resampling
    Sampling and Resampling Thomas Lumley Biostatistics 2005-11-3 Basic Problem (Frequentist) statistics is based on the sampling distribution of statistics: Given a statistic Tn and a true data distribution, what is the distribution of Tn • The mean of samples of size 10 from an Exponential • The median of samples of size 20 from a Cauchy • The difference in Pearson’s coefficient of skewness ((X¯ − m)/s) between two samples of size 40 from a Weibull(3,7). This is easy: you have written programs to do it. In 512-3 you learn how to do it analytically in some cases. But... We really want to know: What is the distribution of Tn in sampling from the true data distribution But we don’t know the true data distribution. Empirical distribution On the other hand, we do have an estimate of the true data distribution. It should look like the sample data distribution. (we write Fn for the sample data distribution and F for the true data distribution). The Glivenko–Cantelli theorem says that the cumulative distribution function of the sample converges uniformly to the true cumulative distribution. We can work out the sampling distribution of Tn(Fn) by simulation, and hope that this is close to that of Tn(F ). Simulating from Fn just involves taking a sample, with replace- ment, from the observed data. This is called the bootstrap. We ∗ write Fn for the data distribution of a resample. Too good to be true? There are obviously some limits to this • It requires large enough samples for Fn to be close to F .
    [Show full text]
  • Efficient Computation of Adjusted P-Values for Resampling-Based Stepdown Multiple Testing
    University of Zurich Department of Economics Working Paper Series ISSN 1664-7041 (print) ISSN 1664-705X (online) Working Paper No. 219 Efficient Computation of Adjusted p-Values for Resampling-Based Stepdown Multiple Testing Joseph P. Romano and Michael Wolf February 2016 Efficient Computation of Adjusted p-Values for Resampling-Based Stepdown Multiple Testing∗ Joseph P. Romano† Michael Wolf Departments of Statistics and Economics Department of Economics Stanford University University of Zurich Stanford, CA 94305, USA 8032 Zurich, Switzerland [email protected] [email protected] February 2016 Abstract There has been a recent interest in reporting p-values adjusted for resampling-based stepdown multiple testing procedures proposed in Romano and Wolf (2005a,b). The original papers only describe how to carry out multiple testing at a fixed significance level. Computing adjusted p-values instead in an efficient manner is not entirely trivial. Therefore, this paper fills an apparent gap by detailing such an algorithm. KEY WORDS: Adjusted p-values; Multiple testing; Resampling; Stepdown procedure. JEL classification code: C12. ∗We thank Henning M¨uller for helpful comments. †Research supported by NSF Grant DMS-1307973. 1 1 Introduction Romano and Wolf (2005a,b) propose resampling-based stepdown multiple testing procedures to control the familywise error rate (FWE); also see Romano et al. (2008, Section 3). The procedures as described are designed to be carried out at a fixed significance level α. Therefore, the result of applying such a procedure to a set of data will be a ‘list’ of binary decisions concerning the individual null hypotheses under study: reject or do not reject a given null hypothesis at the chosen significance level α.
    [Show full text]
  • Resampling Methods: Concepts, Applications, and Justification
    Practical Assessment, Research, and Evaluation Volume 8 Volume 8, 2002-2003 Article 19 2002 Resampling methods: Concepts, Applications, and Justification Chong Ho Yu Follow this and additional works at: https://scholarworks.umass.edu/pare Recommended Citation Yu, Chong Ho (2002) "Resampling methods: Concepts, Applications, and Justification," Practical Assessment, Research, and Evaluation: Vol. 8 , Article 19. DOI: https://doi.org/10.7275/9cms-my97 Available at: https://scholarworks.umass.edu/pare/vol8/iss1/19 This Article is brought to you for free and open access by ScholarWorks@UMass Amherst. It has been accepted for inclusion in Practical Assessment, Research, and Evaluation by an authorized editor of ScholarWorks@UMass Amherst. For more information, please contact [email protected]. Yu: Resampling methods: Concepts, Applications, and Justification A peer-reviewed electronic journal. Copyright is retained by the first or sole author, who grants right of first publication to Practical Assessment, Research & Evaluation. Permission is granted to distribute this article for nonprofit, educational purposes if it is copied in its entirety and the journal is credited. PARE has the right to authorize third party reproduction of this article in print, electronic and database forms. Volume 8, Number 19, September, 2003 ISSN=1531-7714 Resampling methods: Concepts, Applications, and Justification Chong Ho Yu Aries Technology/Cisco Systems Introduction In recent years many emerging statistical analytical tools, such as exploratory data analysis (EDA), data visualization, robust procedures, and resampling methods, have been gaining attention among psychological and educational researchers. However, many researchers tend to embrace traditional statistical methods rather than experimenting with these new techniques, even though the data structure does not meet certain parametric assumptions.
    [Show full text]
  • Monte Carlo Simulation and Resampling
    Monte Carlo Simulation and Resampling Tom Carsey (Instructor) Jeff Harden (TA) ICPSR Summer Course Summer, 2011 | Monte Carlo Simulation and Resampling 1/68 Resampling Resampling methods share many similarities to Monte Carlo simulations { in fact, some refer to resampling methods as a type of Monte Carlo simulation. Resampling methods use a computer to generate a large number of simulated samples. Patterns in these samples are then summarized and analyzed. However, in resampling methods, the simulated samples are drawn from the existing sample of data you have in your hands and NOT from a theoretically defined (researcher defined) DGP. Thus, in resampling methods, the researcher DOES NOT know or control the DGP, but the goal of learning about the DGP remains. | Monte Carlo Simulation and Resampling 2/68 Resampling Principles Begin with the assumption that there is some population DGP that remains unobserved. That DGP produced the one sample of data you have in your hands. Now, draw a new \sample" of data that consists of a different mix of the cases in your original sample. Repeat that many times so you have a lot of new simulated \samples." The fundamental assumption is that all information about the DGP contained in the original sample of data is also contained in the distribution of these simulated samples. If so, then resampling from the one sample you have is equivalent to generating completely new random samples from the population DGP. | Monte Carlo Simulation and Resampling 3/68 Resampling Principles (2) Another way to think about this is that if the sample of data you have in your hands is a reasonable representation of the population, then the distribution of parameter estimates produced from running a model on a series of resampled data sets will provide a good approximation of the distribution of that statistics in the population.
    [Show full text]
  • Resampling: an Improvement of Importance Sampling in Varying Population Size Models Coralie Merle, Raphaël Leblois, François Rousset, Pierre Pudlo
    Resampling: an improvement of Importance Sampling in varying population size models Coralie Merle, Raphaël Leblois, François Rousset, Pierre Pudlo To cite this version: Coralie Merle, Raphaël Leblois, François Rousset, Pierre Pudlo. Resampling: an improvement of Importance Sampling in varying population size models. Theoretical Population Biology, Elsevier, 2017, 114, pp.70 - 87. 10.1016/j.tpb.2016.09.002. hal-01270437 HAL Id: hal-01270437 https://hal.archives-ouvertes.fr/hal-01270437 Submitted on 22 Mar 2016 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Resampling: an improvement of Importance Sampling in varying population size models C. Merle 1,2,3,*, R. Leblois2,3, F. Rousset3,4, and P. Pudlo 1,3,5,y 1Institut Montpelli´erain Alexander Grothendieck UMR CNRS 5149, Universit´ede Montpellier, Place Eug`eneBataillon, 34095 Montpellier, France 2Centre de Biologie pour la Gestion des Populations, Inra, CBGP (UMR INRA / IRD / Cirad / Montpellier SupAgro), Campus International de Baillarguet, 34988 Montferrier sur Lez, France 3Institut de Biologie Computationnelle, Universit´ede Montpellier, 95 rue de la Galera, 34095 Montpellier 4Institut des Sciences de l’Evolution, Universit´ede Montpellier, CNRS, IRD, EPHE, Place Eug`eneBataillon, 34392 Montpellier, France 5Institut de Math´ematiquesde Marseille, I2M UMR 7373, Aix-Marseille Universit´e,CNRS, Centrale Marseille, 39, rue F.
    [Show full text]
  • Profiling Heteroscedasticity in Linear Regression Models
    358 The Canadian Journal of Statistics Vol. 43, No. 3, 2015, Pages 358–377 La revue canadienne de statistique Profiling heteroscedasticity in linear regression models Qian M. ZHOU1*, Peter X.-K. SONG2 and Mary E. THOMPSON3 1Department of Statistics and Actuarial Science, Simon Fraser University, Burnaby, British Columbia, Canada 2Department of Biostatistics, University of Michigan, Ann Arbor, Michigan, U.S.A. 3Department of Statistics and Actuarial Science, University of Waterloo, Waterloo, Ontario, Canada Key words and phrases: Heteroscedasticity; hybrid test; information ratio; linear regression models; model- based estimators; sandwich estimators; screening; weighted least squares. MSC 2010: Primary 62J05; secondary 62J20 Abstract: Diagnostics for heteroscedasticity in linear regression models have been intensively investigated in the literature. However, limited attention has been paid on how to identify covariates associated with heteroscedastic error variances. This problem is critical in correctly modelling the variance structure in weighted least squares estimation, which leads to improved estimation efficiency. We propose covariate- specific statistics based on information ratios formed as comparisons between the model-based and sandwich variance estimators. A two-step diagnostic procedure is established, first to detect heteroscedasticity in error variances, and then to identify covariates the error variance structure might depend on. This proposed method is generalized to accommodate practical complications, such as when covariates associated with the heteroscedastic variances might not be associated with the mean structure of the response variable, or when strong correlation is present amongst covariates. The performance of the proposed method is assessed via a simulation study and is illustrated through a data analysis in which we show the importance of correct identification of covariates associated with the variance structure in estimation and inference.
    [Show full text]
  • Lessons in Biostatistics
    Lessons in biostatistics Resampling methods in Microsoft Excel® for estimating reference intervals Elvar Theodorsson Department of Clinical Chemistry and Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden Corresponding author: [email protected] Abstract Computer- intensive resampling/bootstrap methods are feasible when calculating reference intervals from non-Gaussian or small reference sam- ples. Microsoft Excel® in version 2010 or later includes natural functions, which lend themselves well to this purpose including recommended inter- polation procedures for estimating 2.5 and 97.5 percentiles. The purpose of this paper is to introduce the reader to resampling estimation techniques in general and in using Microsoft Excel® 2010 for the purpo- se of estimating reference intervals in particular. Parametric methods are preferable to resampling methods when the distributions of observations in the reference samples is Gaussian or can tran- sformed to that distribution even when the number of reference samples is less than 120. Resampling methods are appropriate when the distribu- tion of data from the reference samples is non-Gaussian and in case the number of reference individuals and corresponding samples are in the order of 40. At least 500-1000 random samples with replacement should be taken from the results of measurement of the reference samples. Key words: reference interval; resampling method; Microsoft Excel; bootstrap method; biostatistics Received: May 18, 2015 Accepted: July 23, 2015 Description of the task Reference intervals (1,2) are amongst the essential are naturally determined in a ranked set of data tools for interpreting laboratory results. Excellent values if the number of observations is large, e.g.
    [Show full text]
  • Resampling Methods in Paleontology
    RESAMPLING METHODS IN PALEONTOLOGY MICHAŁ KOWALEWSKI Department of Geosciences, Virginia Tech, Blacksburg, VA 24061 and PHIL NOVACK-GOTTSHALL Department of Biology, Benedictine University, 5700 College Road, Lisle, IL 60532 ABSTRACT.—This chapter reviews major types of statistical resampling approaches used in paleontology. They are an increasingly popular alternative to the classic parametric approach because they can approximate behaviors of parameters that are not understood theoretically. The primary goal of most resampling methods is an empirical approximation of a sampling distribution of a statistic of interest, whether simple (mean or standard error) or more complicated (median, kurtosis, or eigenvalue). This chapter focuses on the conceptual and practical aspects of resampling methods that a user is likely to face when designing them, rather than the relevant math- ematical derivations and intricate details of the statistical theory. The chapter reviews the concept of sampling distributions, outlines a generalized methodology for designing resampling methods, summarizes major types of resampling strategies, highlights some commonly used resampling protocols, and addresses various practical decisions involved in designing algorithm details. A particular emphasis has been placed here on bootstrapping, a resampling strategy used extensively in quantitative paleontological analyses, but other resampling techniques are also reviewed in detail. In addition, ad hoc and literature-based case examples are provided to illustrate virtues, limitations, and potential pitfalls of resampling methods. We can formulate bootstrap simulations for almost any methods from a practical, paleontological perspective. conceivable problem. Once we program the computer In largely inductive sciences, such as biology or to carry out the bootstrap replications, we let the com- paleontology, statistical evaluation of empirical data, puter do all the work.
    [Show full text]
  • Resampling Stats Add-In for Excel User's Guide
    Resampling Stats Add-in for Excel User’s Guide Version 4 c statistics.com, LLC 2009 Preface The presentation of resampling methods in this book owes a great debt to Ju- lian Simon—resampling pioneer and creator of the original Resampling Stats software. statistics.com, LLC 612 N. Jackson Street Arlington, Virginia 22201 [email protected] www.resample.com 703-522-2713 i Contents Preface i Contents ii List of Figures v List of Tables x 1 Introduction 1 1.1 How to Use This book ....................... 1 1.2 Installation ............................. 2 1.3 About Resampling ......................... 3 1.4 The Resampling Stats Add-in (“RSXL”) ............. 5 1.5 Probability by Resampling ..................... 6 1.6 Counting Results .......................... 13 1.7 The Frequency Function ...................... 18 2 Advanced Probability 21 2.1 Rates and Results ......................... 21 2.2 Simulation and Hard Problems . 27 3 Confidence Intervals 37 3.1 Confidence Interval for Means ................... 39 3.2 Confidence Interval for a Proportion . 43 3.3 Confidence Intervals for Medians . 45 3.4 Confidence Interval for Profit ................... 47 3.5 Planning Inventory ......................... 49 4 Hypothesis Testing 55 4.1 Resampling and p-values ...................... 55 ii Contents 4.2 Testing for a Difference in Variability . 63 4.3 Resampling in Complex Cases ................... 65 4.4 Multiple Comparisons - Ad Clickthroughs . 71 5 Contingency Tables 79 5.1 Chi-Squared Basics ......................... 79 5.2 Sir Ronald and the Tea Lady ................... 79 5.3 Applying Resampling ........................ 82 6 Correlation and Regression 89 6.1 Applied Correlation: Baseball Salary vs. Rank . 89 6.2 Regression Basics .......................... 92 6.3 Baseball Again: Running Regression from the Resampling Add-in 95 6.4 Multiple Linear Regression: Newspapers and Population .
    [Show full text]