Graduate Program

Statistics Graduate Program

About Virginia Tech Graduate Statistics

Founded in 1949, the Department of Statistics at Virginia Tech is the third oldest in the nation. Our program specializes in training students in statistical theory balanced with extensive applications including practical experience via the Statistical Consulting Laboratory. Over 590 masters degrees and 260 Ph. D. degrees have been awarded by the department. The 18-month Masters program is a model of the time-efficient education of statisticians. The doctoral program includes specialized tracks in traditional and industrial statistics, bioinformatics, and environmetrics.

The department employs 18 full-time faulty to direct approximately 55-60 full-time graduate students. Research areas include bioinformatics, Bayesian theory, design of experiments, environmental risk analysis, nonparametric regression, biomathematics, statistical inference, statistical quality control and improvement, response surface methodology, analysis of directional data, statistical genetics, functional data analysis, and mixed linear and nonlinear models.

To expose students to contemporary topics and to facilitate departmental research, graduate students, undergraduate students, and faculty jointly participate in a series of research and topical seminars. Teaching and research skills are presented to graduate students through a variety of workshops. The department employs a mentoring system where faculty mentor graduate students who, in turn, mentor undergraduate students.

Graduate students are encouraged to participate in internships at companies and industries for the summer (three month internships) or for extended periods of time (up to seven months for Ph. D. students). Course credit is available for a properly monitored and mentored internship experience (Stat 5904 for three to six credits). See Internship in Statistics for more details.

Through the Statistical Consulting Laboratory, students in cooperation with faculty members, become involved in on-campus consulting activities. M.S. students are required to participate in statistical consulting for at least one semester and Ph.D. students for at least three semesters. The department has several laboratories housing state-of-the-art Unix and PC networks. Students have access to these for consulting, course work, and research. Students gain extensive experience with modern statistical software for experimental design, data management and analysis, and computer programming for statistical purposes.

Graduates of the Department of Statistics are now with leading industries, government, and with some of the larger colleges and universities throughout the country and the world. The demand of industry, government, and universities for qualified statisticians exceeds supply, a trend which is expected to continue. In the 2002 Job Related Almanac, statistician is named one of the "ten best jobs" since, "pay is great, the hours are regular, autonomy is high, and job stress minimum." In recent years, 100% of our graduate students obtain employment in statistics upon completion of their degrees.

Additional inquiries about the department should be addressed to:

Dr. Jeffrey B. Birch
Director of Graduate Programs in Statistics
Virginia Polytechnic Institute and State University
Blacksburg, VA 24061-0439
Phone: (540) 231-5630
FAX: (540) 231-3863
E-mail: jbbirch@vt.edu

| Top of Page |

Admissions

Requirements

Admission to the Graduate School is contingent upon receipt of a bachelor's degree from an accredited college or university and upon presentation of evidence of potential to pursue graduate work. Admission normally requires an undergraduate cumulative grade point average of 3.00 or higher on a 4.0 base for the equivalent of the last 2 years of undergraduate study. Exceptions may be made to this rule upon recommendation of the department, providing that the applicant can present other substantial evidence of his or her ability to pursue graduate work at a satisfactory level. In addition, the prospective applicant is required to take the Graduate Record Examination.

The Department of Statistics encourages applications from students in fields other than mathematics and statistics. Although a degree in mathematics is not required, mathematical training through advanced calculus is required. Matrix algebra is desirable but may be taken after enrollment.

Procedures

A complete application packet consists of an application form, two official transcripts of all undergraduate and graduate course work, three letters of recommendation, and a $45 application fee. International applicants must also complete a Financial Certification Form and submit scores on the GRE and TOEFL examinations. All application materials should be addressed directly to the Graduate School.

To ensure efficient processing of the application and consideration for all forms of financial assistance, it is recommended that applicants for regular fall semester admission make certain that their application and all related materials (one copy of the transcript and all test scores) are in the Graduate School and one transcript copy and all letters of recommendation are in the Department of Statistics by January 31. Applications and all related materials for admission or readmission to any other term should reach the Graduate School Office and the Department of Statistics at least eight weeks before the beginning of the term in which enrollment is requested.

It is recommended that all applicants use the on-line application form available at the On-line Admission Form.

Paper application forms and the Graduate Catalog can be obtained by writing to:

The Graduate School
Virginia Polytechnic Institute and State University
Blacksburg, VA 24061-0325

Other Places of Interest to Prospective Graduate Students:

| Top of Page |

Financial

Financial Aid

Graduate assistantships are available with stipends of between $1320 and $1450 per month depending on qualifications at the graduate level. Such appointments allow the appointee to carry a maximum of 12 hours of graduate study per semester. In addition, those on a full assistantship get tuition paid by the University. Research assistantships carrying similar stipends are available for advanced students. Financial aid is available for the summer on a limited basis.

In addition to the regular graduate assistantships (teaching or research), special assistantships are occasionally provided by federal and state agencies.

Students in statistics are also eligible to apply directly to the National Science Foundation for fellowships.

Students desiring to apply for financial aid should so indicate in the application for admission, which should be received before January 31. Awards are announced in February and March. Applications received later will be considered, as long as funds remain available.

Cost

The Instructional Fee (tuition) for full-time students (12 credit hours) for the 2003-2004 academic year is $3,019.50 per semester. This tuition is waived for assistantship holders. All full-time graduate students must also pay a comprehensive fee of $452.50 per semester that includes student health services, the recreational fee, and bus service. For out-of-state students the tuition cost is $4,854.00 per semester, but this is waived for students who earn $4000 or more as a graduate assistants.

To see a complete explanation of tuition and fees click here.

Housing: The University provides a limited number of housing units for graduate students, but many attractive apartments and other housing units are available in Blacksburg and the surrounding area.

| Top of Page |

Masters of Science Degree Requirements

The Department of Statistics offers thesis and non-thesis Master of Science programs, both requiring 35 semester hours, including at least 31 semester hours of work within the department. Students should expect to complete the master's degree in 18 months of graduate study.

The program requires the following first-year core courses:

Fall Inference Fundamentals (3H)
Regression and ANOVA (3H)
Probability and Distribution Theory (3H)
Introduction to Statistical Program Packages (1H)
Spring Statistical Inference (3H)
Experimental Design and Analysis (3H)
Linear Models Theory (3H)
Statistical Consulting (2H)

Additional courses which round out a program may be taken at the graduate level in statistics, mathematics, operations research, or in approved areas of application. Each student will participate for one semester in statistical consulting activities. For graduate assistants, this is usually part of their duties for the assistantship. All Master's students are required to pass two different sections of Special Topics in Statistics (Stat 5984, 1 hour each).

After completing the first-year core courses, each student must pass a written qualifying examination at a level appropriate for the Masters degree. A final oral examination is also required for the M.S. degree.

| Top of Page |

Doctor of Philosophy Degree Requirements

The Ph.D. program offers four concentrations or "tracks", which include the Traditional Track, the Industrial Track, the Bioinformatics Track, and the Environmental Track. The Traditional Track encompasses the general pursuit of research in statistical theory and methods, allowing considerable freedom in choice of coursework within and outside the department. The Industrial, Bioinformatics, and Environmental Tracks offer more specialized statistical training geared towards applications areas in which the department has particular expertise. In accord with their specialized nature, these three tracks are more stringent in requirements for relevant coursework than the Traditional Track. An option in Bioinformatics is also available.

The Ph.D. plan of study requires a minimum of 90 semester hours of work beyond the baccalaureate, including at least 59 semester hours of course work and at least 30 semester hours of research toward the dissertation. In addition to the core courses for the M.S. (or equivalent courses if a student enters the program with advanced standing from another university) required courses for the Ph.D. are Advanced Topics in Statistical Inference and three other Ph. D. level courses from approved lists of courses, which vary by track. Each candidate for the Ph.D. must pass the qualifying examination at the Ph.D. level. The department also offers an option in forestry or wildlife in conjunction with the College of Natural Resources.

While in the Ph.D. program, each student must participate for three semesters in specialized professional training in statistical consulting and/or teaching. For graduate assistants, this will be fulfilled as duty for the assistantship.

In addition to passing the qualifying examination at the Ph.D. level, each doctoral student must pass a written presentation (the "proposal") and oral presentation (the "proposal defense") of their dissertation topic. An oral presentation, in the form of a colloquium, based on the final dissertation results, is also required. The final examination toward the doctorate is the oral defense of dissertation.

Forestry and Wildlife Options: The department offers an option in forestry or wildlife in conjunction with the School of Forestry and Wildlife Resources.

| Top of Page |

Recent Ph.D. Dissertations

  • Mike Joner (2007) Univariate and Multivariate Surveillance Methods for Detecting Increases in Incidence Rates. Chairman: Marion Reynolds and William Woodall

  • Bing Liu (2006) Causal Gene Network Interference from Genetical Genomics Experiments Via Structural Equation Modeling. Chairman: Ina Hoeschele

  • Ying Zhang (2006) Efficient Sampling Plans for Control Charts when Monitoring an Autocorrelated Process. Chairman: Marion Reynolds

  • Li Wang (2006). Recommendations for Design Parameters for Central Composite Designs with Restricted Randomization. Chairman: G. Geoffrey Vining

  • David Farrar (2006). Some Model Based and Nonparametric Clustering Methods for Characterization of Regional Ecological Stressor Response Patterns and Regional Environmental Quality Trends. Chairman: Eric Smith

  • Huizi Zhang (2006). Classification Analysis of Environmental Monitoring: Combining Information Across Multiple Studies. Chairman: Eric Smith

  • Stephanie Pickle (2006). Semiparametric Techniques for Response Surface Methodology. Chairman: Jeffrey B. Birch and Timothy Robinson

  • Xin Zhong (2006). Efficient Sampling Plans for Control Charts when Monitoring an Autocorrelated Process. Chairman: Marion Reynolds

  • Landon Sego (2006). Applications of Control Charts in Medicine and Epidemiology. Chairman: William Woodall

  • Shabnam Modarres-Mousavi (2006). Monitoring Markov Dependent Binary Observations with a Log-Likelihood Base CUSUM Control Chart. Chairman: Marion Reynolds.

  • Willis Jensen (2006). Profile Monitoring for Mixed Model Data. Chairman: Jeffrey B. Birch.

  • Mingjin Yan (2005). Some Methods of Estimating the Number of Clusters and a New Clustering Criterion. Chairman: Ina Hoeschele

  • Penelope Eisenbies (2005). Bayesian Hierarchical Methods and Use of Ecological Thresholds and Changepoints for Habitat Selection Models. Chairman: Eric Smith

  • Yuyan Duan (2005). A Modified Bayesian Power Prior Approach. Chairman: Keying Ye and Eric Smith

  • Younan Chen (2005). Bayesian Modeling on Dual Response Surfaces. Chairman: Keying Ye

  • Peter Parker (2005). Response Surface Design and Analysis in the Presence of Restricted Randomization. Chairman: G. Geoffrey Vining

  • Li Liang (2005). Graphical Tools for Evaluating Split-Plot Designs. Chairman: Christine Anderson-Cook

  • J.D. Williams (2004). Contributions to Profile Monitoring and Multivariate Statistical Process Control. Chairman: Jeffrey B. Birch and William Woodall

  • Mahmoud Mahmoud (2004). The Monitoring of Linear Profiles and the Inertial Properties of Control Charts. Chairman: William H. Woodall

  • Ayca Ozol (2004). Understanding Scaled Prediction Variance Using Graphical Methods for Model Robustness, Measurement Error and Generalized Linear Models for Response Surface Designs. Chairman: Christine M. Anderson-Cook

  • Valentin Parvu (2004). Optimal Blocking for Three Treatments and BIBD Robustness - Two Problems in Design Optimality. Chairman: John P. Morgan

  • Bo Jin (2004). Optimal Designs with Limited Resources. Chairman: John P. Morgan

  • David Lawrence (2003). Cluster Based Bounded Influence Regression. Chairman: Jeffrey B. Birch.

  • Xiao Yang (2003). Optimal Design of Single Factor cDNA Microarray Experiments and Mixed Models for Gene Expression Data. Co-chairmen: Ina Hoeschele and Keying Ye.

  • Ed Boone (2003). Bayesian Methodology for Ecological Data. Chairman: Keying Ye.

| Top of Page |