| | Course Offerings Select a course or scroll
BIOST 2000: Teaching Practicum This course will provide doctoral students with an opportunity to obtain teaching experience. This course is intended for doctoral students during their dissertation stage. Teaching experience will enhance the professional growth of students. Students will further develop oral and written communication skills and an art for explaining material, which is an integral part of a biostatistician's career. Return to top BIOST 2011: Principles of Statistical Reasoning This is the Biostatistics core course for the Graduate School of Public Health. Students obtain an understanding of the concepts of statistical reasoning as applied to the study of public health problems. This includes learning basic terminology and its meaning, the calculations of various statistical measures and indices, quantification of health relationships and the interpretation of inferential statistical techniques. (admission by permission of instructor for Summer Session); Prerequisite: college algebra or a higher-level math course with a grade of C or better Return to top BIOST 2013: Longitudinal Data Analysis The course will cover statistical aspects of analyzing longitudinal data, i.e. data collected on a cohort of individuals, for each of whom the value of a response variable is determined at various points in time. Emphasis will be placed on examples from the biological sciences and medicine. Modeling of both continuous and discrete response variables will addressed. Prerequisite: BIOST 2043 or BIOST 2044 and BIOST 2049 or BIOST 2083; some background in multivariate analysis will be helpful; a working knowledge of linear algebra and calculus is assumed Return to top BIOST 2015: Elements in Statistical Learning The purpose of the course is to present the theory and practice of statistical learning algorithms, placing "statistical learning" or "data mining" techniques in the proper context with regard to their origins in simple classical methods like linear regression, to clarify the strengths and weaknesses from theoretical and practical sides. "Supervised learning" techniques studied include using regularization and Bayesian methods kernel methods, basis function methods, neural networks, support vector machines, additive trees, boosting, bootstrap-based methods. Unsupervised learning techniques studied include cluster analysis self-organizing maps, independent component analysis and projection pursuit. Prerequisites: BIOST 2041, BIOST 2042, BIOST 2043 and BIOST 2044; or permission of instructor Return to top BIOST 2016: Introduction to Sampling Presents practical sampling methods and their theoretical background. Covers simple random, stratified, systematic, and simple stage cluster sampling techniques; also, ratio, regression, and difference estimation. Emphasizes sampling human populations in large communities. Prerequisite: BIOST 2011 or BIOST 2041; or permission of instructor Return to top BIOST 2017: Advanced Sampling Methods Continuation of Biostatistics 2016 (Introduction to Sampling). Focus is on the design and evaluation of complex sample surveys and other observational and experimental studies involving repeated measures and cluster-correlated data. Topics include: the effect of multiple sampling stages, stratification, clustering, unequal probability sampling and non-sampling errors on estimation inference; sample weighting and imputation; capture-recature estimation; adaptive sampling and telephone sampling. Prerequisite: BIOST 2016 and BIOST 2041 and BIOST 2042 and BIOST 2043 and BIOST 2044 and BIOST 2049 and BIOST 2081 and BIOST 2092 Return to top BIOST 2018: Statistical Foundations for Bioinformatics Data Mining Course introduces data analysis methods widely used or rapidly gaining use in bionformatics. Methods deal with prediction, classification, optimization, and clustering and include classification trees, flexible varieties of discriminant analysis including support vector machines. EM algorithm and Monte Carlo Markov chain, bootstrap andbagging, boosting & self-organizing maps. Methods are in context of principles and models of statistical science, with emphasis on Bayesian methods. Examples will be taken from microarrays, analysis of genetic networks, proteomics, computational pharmacology, and research text mining. Prerequisite: three credits of statistics or biostatistics; solid understanding of basic principles of statistics; working knowledge of calculus; and permission of the instructor Return to top BIOST 2019: Public Health Statistics This course reviews the development of biometry and public health statistics and relationships to demography and actuarial science. Introduces students to national health data systems and to life table analysis and measures of fertility, morbidity, and mortality as they apply to public health statistics. Return to top BIOST 2021: Special Studies in Biostatistics Advanced work with approval or guidance of faculty member Return to top BIOST 2025: Biostatistics Seminar Introduces current health problems in biostatistics methods and theory. Includes seminars from faculty at other academic institutions. Return to top BIOST 2030: Social Inequalities in Health The purpose of this course is to introduce students to the current literature in socioeconomic inequalities and their impact on public health. The course will discuss such topics as race, ethnicity and gender as they relate to socio-economic and public health status and critique current methods of uantitative analsyis. Prerequisites: BIOST 2011 or EPIDEM 2110 or HSADM 2000 Return to top BIOST 2035: Experimental Design This course covers a broad perspective of experimental designs covered in public health including various ANOVA designs, case-cohort studies, case-crossover studies, cross sectional studies, prospective and retrospective cohort studies, randomized clinical trials and meta analysis. The advantage and disadvantages of the various studies are discussed and emphasis is placed on selection of the appropriate study, sample size estimation and controlling for sources of bias and reduction of variability. Prerequisites: BIOST 2042 and BIOST 2049 Return to top
BIOST 2040: Elements of Stochastic Processes
Covers generating functions and convolutions of random variables, the Poisson and compound Poisson distributions, ranching processes, random walk, and the gambler's ruin problems, plus Markov chains, and simple birth and death processes. Prerequisites: MATH 0240, BIOST 2034; or permission of instructor Return to top BIOST 2041: Introduction to Statistical Methods 1 This course is an introductory applied biostatistical course for students needing a more comprehensive approach than provided in the Core Course (BIOS 2011). Topics covered include probability, confidence intervals, estimation and hypothesis testing. Statistical procedures covered include t-test, contingency tables, analysis of variance, linear regression and basic nonparametric procedures. Prerequisite: college algebra Return to top BIOST 2042: Introduction to Statistical Methods 2 This course constitutes the second part of the basic sequence of applied statistical methods (components BIOS 2041). The course covers nonparametric methods, multiple linear regression, odds ratios, relative risk, logistic regression, methods in survival analysis (Kaplan-Meier, lifetable analyses, Cox proportional hazards models), multiple comparisons procedures, the general linear contrast, multi-way ANOVA (general fixed effects model, factorial design, qualitative and quantitative interactions, randomized blocks, random effects models, repeated measures) and analysis of covariance. Prerequisite: BIOST 2041; or permission of instructor Return to top BIOST 2043: Introduction to Statistical Theory 1 Basic introduction to statistical theory. Topics covered include joint, marginal and conditional probabilities; moment generating and characteristic functions; transformation of variables; convergence of random variables; law of large numbers; and the central limit theorem. Prerequisite: MATCH 0240; or permission of instructor. Return to top BIOST 2044: Introduction to Statistical Theory 2 Continuation of the introduction to statistical theory introduced in BIOS 2043. Topics covered include sufficiency, completeness, Rao-Cramer's inequality, fundamentals of hypothesis testing, Neyman-Pearson Lemma, and likelihood ratio tests. Prerequisite: BIOST 2043 Return to top BIOST 2045: Analysis of Case-Control Studies Teaches methods to study health problems in community population groups. Covers measures of disease occurrence and association for various study designs; classical analysis of grouped and matched case-control studies. Both conditional and unconditional logistic regressions are covered. Methods are given for analyzing a variable number of controls for each case, assessing the effect of multiple expansive levels, and assessing model fit. Prerequisites: BIOST 2041, 2042; or BIOST 2041, 2042 concurrently; or permission of instructor Return to top BIOST 2046: Analysis of Cohort Studies This introductory applied course in statistical modeling focuses on regression methods for the analysis of cohort data. Topics include the generalized linear model and generalized estimating equations (with emphasis on logistic and Poisson regression), and Cox regression with time-dependent covariates. Students analyze several cohort data sets, assess the adequacy of their models, and interpret their results. Prerequisite: BIOST 2042; and BIOST 2049 Return to top BIOST 2047: Introduction to Biological Assay Teaches statistical techniques in biological assay, including direct and indirect assays, quantitative and quantal responses, estimation of median effective dose, and comparisons of effectiveness. Prerequisites: BIOST 2042, BIOST 2043; or permission of instructor Return to top BIOST 2048: Occupational Biostatistics Topics covered include nested case control studies, design and analysis of historical prospective studies, the statistical package OCMAP, evaluation of disease clusters, and the multistage model. The course considers the practical problems of exposure assessment, data collection, and adjustment for confounders in addition to the selection of the appropriate statistical method. Prerequisite: BIOST 2042 Return to top BIOST 2049: Applied Regression Analysis Covers the basics of classical and modern regression techniques. Topics covered include multiple regression, indicator variables, multicolinearity, selection of a best model, influence diagnostics, and nonlinear regression. Prerequisite: BIOST 2042; or permission of instructor Return to top BIOST 2051: Statistical Estimation Theory This course covers selected topics in classical estimation and builds on some of the estimation topics covered in Biostatistics 2044. Topics covered include comparisons of different methods of estimation, properties of maximum likelihood estimations, brief overview of measure theory, and the decision theoretic framework work for estimation. Although this course is primarily theoretical practical issues are discussed such as convergence problems when obtaining ,maximum likelihood estimations, impact of missing data, and the role of some of the recent computationally intensive methods. Prerequisite: BIOST 2044; or permission of instructor Return to top BIOST 2052: Multivariate Analysis Topics covered include the multivariate normal distribution, estimation of the mean vector and covariance matrix, distributions and uses of simple, partial and multiple conclation correlation coefficients, the generalized T2 statistic, the distribution of the sample generalized variance, multivariate analysis of variance and the multivariate Behrens-Fisher problem. Multivariate methods are applied to repeated measures analysis, factor analysis, and discriminant analysis. The beginning of the course emphasizes theory. Later applications and computational methods are emphasized. Several lectures are devoted to the review and presentation of current and classical literature involving methods in multivariate analysis. Prerequisite: BIOST 2044; or permission of instructor Return to top BIOST 2053: Nonparametric Methods in Statistics Order statistics and quantiles: the U statistic and Hoeffding's theorem; ranks and mid-ranks; Kendall's and Spearman's rank correlation coefficients; the sign test; Kolmogorov-Smirnov and Cramer-Von Mises tests of goodness of fit; rank (order) test; Mann-Whitney (Wilcoxon) rank test; the several-sample problem; the Mood Brown median test; Mood's median test; the Kruska-Wallis H test; Friedman's test; efficiency of nonparametric tests; Pitman efficiency of tests. Prerequisite: BIOST 2044; or permission of instructor Return to top BIOST 2054: Survival Analysis Covers the basic theoretical aspects of various models to analyze "time to event" data. Introduces basic concepts such as the survival function, hazard function, left and right hand censoring, and common parametric models for analyzing survival data. Also includes the proportional hazards model with fixed and time dependent covariates, the stratified PH model, regression diagnostics for survival models, additive hazards regression models and multivariate survival models. Prerequisite: BIOST 2044; or permission of instructor Return to top BIOST 2055: Statistical Methods & Data Mining in Micro-array Analysis In this course, we will discuss statistical methods and data mining techniques used in modern genomic experimental data. Special emphasis will be on microarray expression profiles though other related topics including protein-protein interaction and mass spectrometry data will also be touched. Some of the methods are typical traditional methods but many others are rapidly evolving which may not be taught in any standard statistical course. We will discuss statistical issues in data processing in various microarray platform. Gene screening, visualization, clustering and classification of microarray data will be emphasized. Methods of detecting genetic regulatory network will also be touched. R and Bioconductor will be mainly used as the software tool in class while many freely available packages will be introduced in hands-on labs. Prerequisite: One year of elementary statistical course (BIOST 2041 and BIOST 2042 or equivalent). Return to top BIOST 2061: Likelihood Theory & Applications The purpose of this course is to introduce the student to modern likelihood theory and its applications. The course will cover maximum likelihood theory, profile likelihood theory, pseudo likelihood theory and generalized estimating equations. The course is taught at the doctoral level and much of the theory is illustrated through applications. Prerequisites: BIOST 2043 and BIOST 2044 Return to top BIOST 2062: Clinical Trials: Methods and Practice Course consists of two weekly lectures, posted on the web in advance, and two in-class sessions which consist of questions and answers related to the web-based information, problem-solving, or discussion of case studies. It covers fundamental concepts in the design and conduct of modern clinical trials. Topics include: experimental designs for safety and efficacy trials, quantitative methods for design, interim monitoring, and analysis of randomized comparative clinical trials including crossover, factorial and equivalence designs. Ethical, organizational, and practical considerations of design and conduct of single and multicenter studies are integrated in lectures and case studies. The course also covers international guidelines on statistical considerations for drug development, guidelines adopted for publication of trials in major medical journals, and recommended approaches for meta-analyses. Prerequisites: The equivalent of a one-term course in introductory biostatistics is recommended. Permission of instructor. Return to top BIOST 2063: Bayesian and Empirical Bayes Statistical Methods Teaches how to use Bayesian and empirical Bayes statistical methods in data analysis. Gives examples and includes discussion of common data-analysis issues. Compares and contrasts Bayesian, empirical Bayesian, and classical methods. Topics covered include Savage axioms, conjugate priors, the likelihood principle, noninformative priors, Empirical Bayes and hierarchal Bayes approaches. Prerequisites: BIOST 2042, BIOST 2044 Return to top BIOST 2064: Bayesian and Empirical Bayes Computational Methods This course provides the students with an understanding of both the theory and practice with regard to the EM algorithm, Markov-chain, sampling techniques, importance sampling, and the solution of decision trees. Students gain hands-on experience programming with S-Plus. The course is designed to complement BIOS 2063. Prerequisite: BIOST 2063 Return to top BIOST 2065: Analysis of Incomplete Data This course will introduce the missing-data problems and implication in statistical inference. Naive methods, standard likelihood-based methods, theory and application of multiple imputation, data augmentation, Gibbs sampler, and some robust statistical methods will be discussed. Need to program for the homework assignments or final project. Prerequisite: Calculus, linear algebra, and elementary probability; BIOST 2043, BIOST 2044 and BIOST 2051; knowledge in categorical data anlaysis or generalize linear models is recommended.
Return to top BIOST 2081: Mathematical Methods for Statistics This course is designed to bridge the gap between students with an undergraduate mathematics background and the mathematics required for the mathematical statistics courses. It is recommended for students who meet only the minimum requirement of one year of calculus at the time of entry into the program. It covers selected topics in advanced calculus and linear algebra. Prerequisite: 1 year of college calculus Return to top BIOST 2083: Linear Models Teaches linear model techniques for analyzing balanced and unbalanced data. Basic topics covered include properties of quadratic forms, noncentral chi-square and F distributions, best linear unbiased estimations, and likelihood ratio test. Also covers generalized inverses, models not of full rank, orthogonal contrasts with unbalanced data, regression on dummy variables, analysis of covariance; and analysis of variance components. Prerequisites: BIOST 2044, BIOST 2081; or permission of instructor Return to top BIOST 2084: Discrete Multivariate Analysis This more advanced course on modeling multivariate categorical data focuses on the theory and methods underlying both asymptotic and exact inference. Log-linear models, models for ordinal and multinomial data, exact logistic regression, and extensions of generalized linear models to correlated discrete data are emphasized. Prerequisite: BIOST 2044; and BIOST 2046 Return to top BIOST 2085: Applied Time Series Analysis Covers time series, estimation in the time domain, forecasting, and spectral analysis. Stresses computer application with real data sets. Prerequisites: BIOST 2043, an introductory course in computing; or permission of instructor Return to top BIOST 2086: Applied Mixed Models Analysis Mixed model analysis provides a new approach to modeling which allows one to relax the usual independence assumptiosn and take nito account comlicated data structures. This course will consider all types of mixed models into a general framework and consider the practical implications of their use. Topics will include; normal mixed models, generalized mixed models, and mixed models for categorical data, repeated measures data analysis and cross-over trials with mixed models. Software for fitting mixed models will be discussed. Prerequisites: BIOST 2083; or permission of instructor Return to top BIOST 2087: Biostatistics Consulting Practicum For second-year master's and doctoral students. Provides exposure and experience in consulting on the biostatistical aspects of research problems in biomedical or allied fields. Students have discussions with investigators, leading to the design or analysis of a current research problem. Includes weekly group discussions with instructor. Prerequisite: admission by permission of instructor Return to top Special Topics 0-8: 2070(2), 2071(2), 2072 (2), 2075(3), 2076(3), BIOST 2088(1), 2089(1), 2090(3), 2091(1) Introduces the student to specialized topics in biostatistics in areas of current interest to the field. Two to four special topics courses are given each year. These courses are given to broaden the exposure of the students to the large variety of existing methods and applications in biostatistics. Courses may be one, two or three credits and may be given by faculty with secondary appointments and visiting faculty as well as faculty with primary appointments in the Department of Biostatistics. Special topics given in recent years include: meta analysis, multiple comparisons procedure in clinical trials, methods and applications with ROC curves, special topics in mixed models, statistical methods in AIDS research and smoothing techniques using splines. Prerequisite: variable, depending on specific topic; permission of instructor may be required for admission Return to top BIOST 2092: Introduction to Computing Provides an overview of the University of Pittsburgh computing and electronic resources; An introduction to MINITAB, an interactive statistical software package for storing, displaying, summarizing and analyzing data. Both the command line and Windows menu approaches of the program will be presented; Techniques for computerized bibliographic database searching and the use of electronic library resources for public health. Return to top BIOST 2093: Data Management and Analysis Students obtain a working knowledge of two statistical analysis software packages, SAS and SPSS. Emphasis will be placed on the basics of data management of files, data manipulation, basic data display, descriptive statistics, frequency distributions, and graphical display of data. Although the Windows environment will be discussed, emphasis will be placed on the writing of program code. Return to top BIOST 2095: Indroduction to Database Management Systems Students will obtain an understanding of database models with a specific working knowledge of databases used on the PC. Microsoft Access running under Windows will be used to exemplify the concepts, techniques and examples. Students will learn to design, create and modify databases. In addition, they will learn how to query the database, generate reports, create subsets of the data, and import and export files. The basics of database applications design and development will be presented. The concepts presented are universal and applicable to other database management systems. Return to top BIOST 2096: Numerical Methods in Biostatistics The purpose of this course is to familiarize students with a broader range of numerical methods which are useful in biostatistical research. Selected computational techniques used in statistical research are covered. Some background will be provided to facilitate understanding of a few numerical algorithms widely used in statistics. The following topics are covered: recurrence relations, power series and asymptotic expansions, generating pseudo-random deviates, basic simulation methodology, solutions of nonlinear equations, Newton's method, vector and matrix norms, linear regression and matrix inversion, independent Monte Carlo and Markov Chain Monte Carlo methods. Prerequisites: BIOST 2043 and BIOST 2044 and BIOST 2049; or permission of instructor Return to top BIOST 3010: Research and Dissertation for the Doctoral Degree Variable number of credits assigned for work on student research dissertation. Return to top BIOST 3023: Geographic Information Systems & Spatial Data Analysis This course covers the use of GIS and spatial data analysis techniques in empirical public health research. Basic descriptive and analytical functions of GIS are introduced along with spatial and geographic concepts. The interrelationship between GIS and spatial data analysis will be demonstrated through the use of specialized GIS and spatial data analysis software with a particular emphasis on the study of spatial patterns and spatial autocorrelation in public health research. Return to top FTDR 0000: Full-time Dissertation Study For doctoral candidates who have finished all credit requirements and minimum dissertation requirements. Full-time work on dissertation. No credit. |  |
