Masters in Applied Statistics Overview The Master of Science in Applied Statistics (MAS) degree is designed for professionals or students with undergraduate degrees in the sciences or business. The MAS program will train student in applied statistical methods used in industry and government for process and quality improvement. A key focus of the training is the continuous training and practice using Six Sigma methodology of process improvement. These techniques are in demand in industry due to the cost savings typically resulting for process improvement activities. Use of statistical computer packages will be part of all applied courses. At the end of the program, students will understand how to: 1) use many applied statistical tools, 2) analyze data using several statistical computer packages, 3) identify opportunities for quality and productivity improvement, 4) establish process control systems, 5) lead problem solving teams, and 6) communicate orally and in writing project results. The 36 hour program is scheduled to enable completion in 22 months (5 semesters including one summer). The MAS program is unique from traditional statistics degree programs in the following areas: Paired Block of Courses – Each semester a course is offered in a “Methods” block and an “Applied” block. The content of the two blocks of study will be coordinated to enable the student to gain an understanding of underlying statistical methods at the same time they are applying related applied statistical tools. This Methods/Applied format also provides the students with varied material each semester making the learning experience more enjoyable. Building Analysis Capability Each Semester – The paired block design provides the student increasing capability to analyze problems with each successive semester. The student does not wait until after several semesters to gain a capacity to perform useful analyses. Even after completing one semester, the student is able to perform practical quality and process improvement studies. Should the coursework be interrupted, the student will still be left with practical, useful skills. 3/9/2016 1 Statistical Computing – Starting the first semester the student will utilize statistical programs such as SAS, JMP and Minitab to analyze data and present graphical summaries. Students will be taught how to build data bases using programs such as Excel. The use of statistical computing tools will continue for all courses. The skills learned in this area will enable the student to be effective in the workplace early in the program. Applications Project – Students will be required to participate in a one-hour credit project activity for each semester. This project will involve the analysis of data from their workplace where possible. Students will analyze results with direct supervision and present results at a seminar with faculty and peers attending. Both technical and presentation skills will be the focus of the training and evaluation. The final semester will be a two-hour project course that will require the student to make both an oral and written presentation. Nontraditional Schedule – For each of 5 semesters, a 3 hour Methods course, 3 hour Applied course, and a one-hour project will be offered (two hours in the final semester). Typically, courses would be offered two days a week at 5:00PM and 6:30PM enabling both traditional and nontraditional students to participate in the program. Boot Camp Option – The summer prior to the start of the program students will have an option of taking a refresher course in calculus that will focus on the methodology needed to be successful in courses in the Methods block. Also, a computer course will be offered for students needing a refresher in computer skills required for courses in the Applied block. Electives – Two courses in the Applied block can be substituted for graduate courses in Information Systems (Information Systems Project Management Methods - IS8050) or Economics (Econometrics and Forecasting Methods – ECON 8700 or Multivariate Data Analysis – ECON 8720). Ph.D. Program Transfer Ability – The Methods courses mirror many traditional master degree programs. Upon completion of the MAS program the student may elect to transfer to a research university offering a Ph.D. degree. Depending on the university, the MAS program may lead directly into a Statistics Ph.D. program. To help prepare the students for their selected research university, the student may substitute two courses in the Applied block for directed studies in such areas as measure theory-related probability studies. Depending on the student’s area of interest, the 2 elective courses can be selected to prepare for a Ph.D. program. 3/9/2016 2 Six Sigma Training - The final course is Six Sigma and Problem Solving Methods. All courses in the Applied block will reference material to Six Sigma methodology. The final course will pull together all techniques and the application of Six Sigma to process improvement methodology. The Black Belt practice exam will be reviewed to help prepare student to take the exam should they choose. Combined with the project requirement, students should be able to meet Black Belt requirements, assuming proper workplace support. Careers The entry degree for most positions requiring statistical training is the masters degree. A recent Bureau of Labor Statistics report indicated that 18% of the country’s statisticians work for the federal government, 16% for state and local governments and the remainder for private industry. University based statisticians are a relatively small percentage. Thus, a large percentage of MAS degree students will likely be placed in the private sector. From Amstat News (Dec 2003 pp. 8-13) the average starting salary in 2003 for an MS statistician ranged between $45,000 in federal government to $60,000 in the pharmaceutical industry. An August 2004 web search of American’s Job Bank for positions in the Atlanta area with a Six Sigma keyword produced 96 jobs from a variety of companies; most positions were for over $50,000 per year. Applied statisticians are used in industry and government where skills are needed for areas such as: testing, sampling surveys, computing, market research, reliability, estimation, quality control, and process improvement. Some area that employ statisticians include: manufacturing, medicine, pharmacology, public health, engineering, consulting, insurance, defense, economics. The American Statistical Association web site (www.amstat.org/careers) gives an extensive list. 3/9/2016 3 The unique feature of the KSU program in applied statistics is the combination of a coordinated methods/applied course paring combined with a focus on learning the Six Sigma approach to process improvement. Students completing the MAS program will have more statistical training than the typical Six Sigma Black Belt recognized by industry as a desirable process improvement credential. The reason KSU is able to offer such a relevant, in demand program is due to its unique faculty in the Department of Mathematics that university and industry experience. Faculty A complete list of faculty with background information can be obtained of the KSU Math Department web site (www.math.kennesay.edu). A brief background of faculty participating in MAS teaching appears below: Victor Kane – Department Chair, Ph.D. in statistics from Florida State University and MBA in Management from University of Tennessee. Dr. Kane has recently joint KSU from a 20 year career with Ford Motor Company. He has served as a Six Sigma Champion where he was a Plant Manager. His career included both technical and management assignments with Ford. He is the author of a process improvement book, Defect Prevention. Marla Bell – Assistant Department Chair, Associate Professor of Mathematics Ph.D. Clemson University. Dr. Bell's Master's thesis was in Applied Statistics. Her PhD dissertation was in Queuing Theory. She came to Kennesaw State in 1994. During her time at KSU, she has been most interested in effective statistics education. Most recently she authored a calculator guide to supplement a best selling Elementary Statistics book. She is also active in statistical consulting. Anda Gadidov – Assistant Professor of Mathematics, PhD in Mathematics, Texas A&M University. Prior to joining KSU in 2004, Dr. Gadidov taught in Pennsylvania, and at Georgia Institute of Technology. Her research focuses on probability theory, with emphasis on U-statistics, and applications of probability and statistics in finance. Dr. Gadidov is the author of several papers which appeared (or will appear) in well known journals, like The Annals of Probability and Statistics and Probability Letters. 3/9/2016 4 Jennifer Priestley - Assistant Professor of Mathematics, Ph.D. in Decision Sciences from Georgia State University, an MBA with a concentration in Quantitative Business Analysis from The Pennsylvania State University and a BS in Economics from Georgia Tech. Dr. Priestley recently joined KSU after 12 years in management consulting and the financial services industry, where she worked for Andersen Consulting, MasterCard, AT&T and VISA EU. She has authored several papers on statistical modeling and data analysis. Lewis VanBrackle – Professor of Mathematics, Ph.D. in statistics fromVirginia Polytechnic Institute and State University. Dr. VanBrackle joined KSU in 1984 after working as an oil-field engineer in South America, a member of the technical staff at Bell Telephone Laboratories and an Associate Staff Manager at BellSouth. He has been a Visiting Scientist at the Centers for Disease Control and Prevention. He has co-authored papers in areas as diverse as epidemiology, computer science education and quality control. 3/9/2016 5 Course Schedule The following is a typical program schedule: Semester Methods Block Summer Calculus Refresher Fall Distribution Theory & Inference (3) Spring Applied Block Project Hours Computer Refresher Quality Control & 1 Process Improvement (3) 7 Analysis of Variance & Regression (3) Statistical Computing & 1 Simulation (3) 7 Summer Mathematical Statistics (3) Measurement Systems Analysis (3) 1 7 Fall Design of Experiments (3) Categorical Data Analysis (3) 1 7 Spring Applied Multivariate Methods (3) 1 8 _______ Six Sigma and Problem Solving (3) Total 3/9/2016 6 36 Brief Course Descriptions Distribution Theory and Inference: Fundamental concepts of probability, random variables and their distributions; review of sampling distributions; theory and methods of point estimation and hypothesis testing, interval estimation, nonparametric tests, introduction to linear models. Statistical software packages such as JMP, MINITAB and/or SAS will be used. Text: Introduction to Mathematical Statistics by Hogg and Craig. Mathematical Statistics: Point estimation, method of moments, maximum likelihood, and properties of point estimators; confidence intervals and hypothesis testing; sufficient statistics; Neyman-Pearson theorem, uniformly most powerful tests, and likelihood ratio tests; Fisher information and the Cramer-Rao inequality. Additional topics may include nonparametric statistics, decision theory and linear models. Text: Introduction to Mathematical Statistics by Hogg, McKean and Craig. Statistical Inference by Casella and Berger. Applied Multivariate Methods: Relevant matrix theory, multivariate random vectors, exact and asymptotic distributions, multivariate normal distribution (MVN), Q-Q plots, sampling from MVN and inference for population mean vector, covariance matrix, correlation matrix, MANOVA, principal component analysis, factor analysis, discriminant analysis, classification and clustering. Statistical software packages such as JMP, MINITAB and/or SAS will be used. Text: Applied Multivariate Statistical Analysis by Johnson and Wichern. Analysis of Variance and Regression: one-way analysis of variance, multiple comparison procedures, contrasts, random and fixed effect models, transformations, experimental design, nested designs, randomized blocks, factorials, nonparametric methods for one and two-way analysis of variance, simple and multiple regression, correlation, diagnostics for detection of multicollinearity analysis of covariance. Statistical software packages such as JMP, MINITAB and/or SAS will be used. Text Applied Linear Regression Models by Neter, Kutner, Wasserman and Nachtsheim. Design of Experiments: Methods for constructing and analyzing designed experiments are considered. Experimental unit, randomization, blocking, replication, and orthogonal contrasts; error reduction and treatment structure, completely randomized design, randomized complete block design, incomplete block designs, Latin squares design, splitplot design, repeated measures design, factorial and fractional factorial designs. Statistical software packages such as JMP, MINITAB and/or SAS will be used. Text: Design and Analysis of Experiments by Montgomery. Statistics for Experimenters by Box, Hunter and Hunter. 3/9/2016 7 Quality Control & Process Improvement: Classical quality control methods such as control charts and sampling plans will be integrated with process improvement tools such as process flowcharts and simple graphical statistical tools. Utilization of standard computer packages for data analysis will be emphasized. Texts: Statistical Quality Control by Grant and Leavenworth. Defect Prevention by Kane. Statistical Computing & Simulation: Foundational topics in Simulation modeling such as model development/design/execution, discrete versus continuous simulation, selecting and fitting different input distributions, analysis and interpretation of model output and variance reduction methods will be addressed through case studies and real-world examples. The course will heavily utilize standard computer packages. Text: Applied Simulation Modeling by Seila, Ceric and Tadikamalla. Measurement System Analysis: Discrete and continuous data collection including operational definitions, stratification and Measurement System Analysis (MSA) will be presented. Methods to assess accuracy and precision of the measurement systems will be studied. Texts: Measurement System Analysis AIAG Action Group, Gauge R&R Studies by Perez-Wilson Applied Categorical Analysis: Methods of contingency table analysis, generalized linear models, logit and log-linear models, logistic and Poisson regression. Applications to Statistical software packages such as JMP, MINITAB and/or SAS will be used. Text: An Introduction to Categorical Data Analysis by Agresti. Six Sigma Problem Solving: The focus of this course is applying Six Sigma methods such as DMAIC to industrial problems using the statistical methods studied in prior courses. Students will analyze industrial data and debate appropriate approaches to utilizing Six Sigma methods. Since Six Sigma method will be utilized throughout the program this course is a synthesis of prior learning. Student will take the American Society for Quality practice Black Belt exam to help prepare them for the exam. The class will review exam questions and address areas where students are having difficulty. Text: The Six Sigma Way by Pande, Neuman and Cavanagh. Note: The above texts may be changed. Courses will be offered at the level of the indicated texts. 3/9/2016 8