Franklin Arno Graybill, Professor Emeritus and founding Chair of the Department of Statistics, died Friday, February 17, 2012. He is survived by his wife Jeanne, his children Dan and Kathy, three grandchildren, and four great-grandchildren.
Frank wrote his dissertation under the direction of Professor Oscar Kempthorne, graduating with a PhD from Iowa State University in 1952. He served on the faculty of Oklahoma A&M until 1960, when he began his tenure at Colorado State University. During his initial years at Colorado State University, Frank created the Statistical Laboratory (on January 13, 1961) and also initiated the PhD program in Statistics. He successfully recruited and hired many talented faculty members. In 1971, the Department of Mathematics and Statistics split into two separate departments and Frank became the head of the newly born Department of Statistics. He served as department head until 1975.
Frank was the editor of Biometrics during 1972–1975, and was the president of the American Statistical Association in 1976. He advised 33 MS and 26 PhD students during his tenure at CSU, and published a number of books, both theoretical and applied.
The success of the statistics program at Colorado State University is in a large part due to Frank’s efforts, his astute hiring during the 1960s, his vision for the department, his creation of the statistical laboratory and his leadership and hard work over the years.
Frank's legacy is exemplified by the fact that we named our consulting unit after him: Graybill Statistical Laboratory, as well as our annual conference series, Graybill Conference, which has been on-going since 2001. We also have the endowed Graybill Award for Excellence in Linear Models, awarded to a graduate student each year.
To continue honoring Frank Graybill's legacy at CSU, the Department launched an initiative to create a $50,000 endowment for the Graybill Conference series, with the goal of ensure that we can continue hosting these conferences, attract top speakers and pay for students to attend. To make a donation towards this endowment, see https://advancing.colostate.edu/GRAYBILL_CONFERENCE.
For a more comprehensive look at Frank Graybill’s professional life and achievements, see http://www.stat.colostate.edu/statdepartment/statnews/stat2012graybillobit
Dear Alumni and Friends of the Department of Statistics,
Welcome to the new departmental newsletter. As you can see, it is in a completely revamped format, which should make it easier to read on the increasingly wide range of electronic devices we have access to nowadays. We hope that you will like it, and we welcome your feedback. A lot of events happened in the department in 2012, some good and some sad. I will highlight some of them, but I encourage you to read the stories below as well.
As most of you know, Frank Graybill passed away last February, at the age of 90. He will be missed by many CSU employees and alumni, who knew him as advisor, mentor, teacher and friend. I only came to CSU in 2007, so I never met Frank in person, but as a member of the Department of Statistics, I am in contact every day with his professional legacy; this is the department he created, one of the top statistics departments in the country. Frank's legacy is exemplified by the fact that we named our consulting unit after him, the Graybill Statistical Laboratory, as well as our conference series, the Graybill Conference. We also have the endowed Graybill Award for Excellence in Linear Models, awarded to a graduate student each year. With the generosity of many of you, we are in the process of creating an endowment for the Graybill Conference series, which will allow us to continue holding these popular events each year.
More recently, Rick Gumina, a long-time instructor in the department, died suddenly on October 22. He was 55 years old. Rick had been teaching undergraduate statistics courses since 1997 and obtained his MS in Statistics in 1998. Many of our graduate student alumni since that time will remember Rick, because in addition to teaching these courses, he was responsible for coordinating the recitations that were led by graduate students. I’ve already heard from several alumni who told me that Rick is the person who “taught them to teach” statistics. He will be missed by the current CSU employees and students who interacted with him, and by the thousands of students he taught over the years. A short obituary is available here.
There have been a number of exciting new changes in the department this year. At the graduate level, we started a completely new degree, the Master of Applied Statistics (MAS). The separate story in this newsletter describes some of its unique features, including its intensive one-year curriculum comprised of eight-week “sub-terms” and the capstone consulting class in June. This degree is intended to meet the growing demand for statistical analysts in government, industry and academia and has generated a lot of interest already among prospective students and employers. Closely associated with the launch of new MAS degree, we are completely revamping the Distance offerings of the department; over the next year, we will be phasing out the Distance MS and replacing it by the Distance MAS, which students will be able to complete in as little as two years. We believe that this degree will better correspond to the interests of the Distance students, who are often already working as data analysts and want to learn the tools to be more effective in that role.
At the undergraduate level, there has been a steady increase in the number of students seeking degrees in statistics, which currently include the Concentration in Statistics within the Major in Mathematics, and the Statistics and Applied Statistics Minors. But the most exciting news is not so much what happened, but what is being planned for next year; we have started to process to reintroduce the Major in Statistics, which was eliminated in 2001 and converted to the Concentration mentioned above. This will replace the Concentration, giving us once again full control over that degree, enhancing its visibility and differentiating it more clearly from Mathematics.
As we emerge from the difficult economic times that resulted in substantial budget reductions at the university and departmental level, we are excited to be entering a period of growth for the department. With strong support from the College of Natural Sciences, we have started three faculty searches this year, two of which are new positions. We expect to be able to add further new positions in the next few years as well. This, combined with the continued growth of statistics as a discipline, makes me very optimistic about the future of the Department of Statistics at Colorado State University.
Finally, let me announce the next Graybill Conference, scheduled for June 10-12, 2013 in Fort Collins. The topic will be survey statistics and is co-organized by Jay Breidt and myself. We hope to see many of you there.
I would be delighted to hear from you any time, whether it is because you have a question about the department or just want to update us on what is happening in your life. As you will see from the story about Sunny Bak Hospital’s trip to North Korea in this newsletter, we would love to share stories, big or small, about our alumni in future issues. And of course, if you are ever in Northern Colorado, do not hesitate to visit us on the Oval!
On June 17-20, the department held the 10th Graybill Conference. These conferences have been held (almost) every year since 2001, on a wide range of topics. This year, the conference was a joint event, because it also served as the annual meeting of the Western North American Region of the International Biometric Society (WNAR). The WNAR annual meetings are hosted each year by a university in its target region of Western North American, but this was the first time it came to Colorado State University. The conference was very successful, with almost 200 attendees and 100 talks.
The Graybill Conference was named after Professor Franklin Graybill, to honor his legacy to the department and the discipline. As announced elsewhere in this newsletter, Frank Graybill passed away in February, so a number of special events were added to the conference, including a plenary talk by Dr. Richard Burdick entitled "Franklin Graybill's Contributions to Variance Component Models with Applications" and a special memorial event for his family and friends.
As announced elsewhere in this newsletter as well, the Department is in the process of creating an endowment for the Graybill Conference series. See https://advancing.colostate.edu/GRAYBILL_CONFERENCE if you wish to contribute to this endowment.
Picture: (right) Professor Frank Graybill in his office
Funded by a $1.1 million, four-year grant from the NSF/NIGMS program run jointly by the National Science Foundation and the National Institutes of Health, a team of statisticians, mathematicians, biochemists, and computer scientists at Colorado State University is devising solutions to the problem of identifying complex biological molecules by examining images obtained from scattering X-rays. Jay Breidt, Professor of Statistics, serves as the principal investigator of the project. Don Estep, Professor of Statistics and Mathematics, is a co-investigator.
Determining the shape of a new molecule is fundamentally important in discovering how the molecule functions with respect to human health and disease, but determining the shape is an “ill-posed” inverse problem that presents significant challenges for computer science, mathematics, and statistics. The difficulty of the problem can be understood by imagining a single, time-lapse photo of an entire basketball game: all of the players running, dribbling, passing, and shooting. Now imagine identifying a basketball player from that one blurry blob of a photograph. Replace the photo by an image collected from a high-intensity X-ray and replace the players by complex biological molecules floating in solution, and that is the problem of “Small-Angle X-ray Scattering” or SAXS. The X-rays scattered from the molecules are collected on a detector analogous to a digital camera. These detector data are used to fit scattering intensity curves. Inverse Fourier transforms of the intensity curves lead to inter-atomic distance distribution functions, and these functions are used to identify molecular structures.
The project team is developing new computational methods and analyzing the effects of data and modeling error on identified structures. Results so far include new models for the autocorrelation structure inherent in SAXS images, and new, statistically rigorous methods for the optimal estimation of molecular features.
Summer 2012 launched our new Master’s of Applied Statistics program. The MAS is a one-year, professional-type degree. Compared to our “regular” MS, the MAS program emphasizes practical methods in statistics and de-emphasizes the theoretical development. The MAS curriculum is entirely separate from that of the MS; it covers more topics, but in less theoretical depth.
The program comprises 19 courses and 31 credits, and can be completed in 11 months. The program can be taken on-campus or via Distance. The on-campus program begins in the three weeks prior to the start of the fall semester with an intensive math review and an introductory computing course. Each of the fall and spring semesters is divided into two eight-week sub-terms. In each of these sub-terms, the students take three two-credit classes and a single one-credit class. The program is completed with a capstone consulting class in May/June.
For those students interested in the earning this degree via Distance, the courses can be taken in two or more years. The courses are the same as taken on campus and they are fully transferable between the two modes of delivery.
The objectives of the new MAS program are to provide rigorous training for applied statisticians, to provide statistical training for students and professionals in other disciplines, and to provide an opportunity for a fifth-year masters degree for CSU undergraduates with statistics or actuarial concentrations.
The MAS program is intended for students with undergraduate backgrounds in math, science, social science, or business fields who see the need for additional education and training but do not want to pursue a research career path. In addition, we anticipate interest from PhD students in other fields for whom training in statistical methods would be beneficial. The MAS classes represent a substantial increase in the number of graduate-level service courses offered by the department. The pre-requisites are: an undergraduate degree with three semesters of calculus, a course in linear algebra, and at least one undergraduate-level statistics course.
The MAS degree is consistent with our recent strategic planning, which includes the improvement of service and core courses in statistics at the graduate level, and strengthening, modernizing, and expanding our distance degree program. It simultaneously accomplishes these strategic goals, by providing a professional degree to serve a type of student not suited for, or not desiring, either the standard MS or the PhD. The MAS program will put CSU at the forefront of graduate education nation-wide in a currently underserved area.
For more information about the MAS program, visit the departmental website at http://www.stat.colostate.edu/statprostudents/statmas.html.
Picture: On-campus MAS students from the class of 2013. (Left to Right) Harold Gomes, Rebecca Clark, Teng Liu, Seth Johnson, Kylie Stitt, Emily Schroder, (front) Sean Barnes
Undergraduate Enrollment in Statistics at All-Time High
Our undergraduate statistics concentration is growing, with 31 majors (an all-time high) and 18 minors. This was perhaps helped by Google’s recognition of the importance of statistics. Hal Varian, Google’s Chief Economist, recently said, “I keep saying the sexy job in the next ten years will be statisticians. People think I'm joking, but who would've guessed that computer engineers would've been the sexy job of the 1990s?” (The McKinsey Quarterly, Jan 2009). (For more from Mr Varian, see www.youtube.com/watch?v=D4FQsYTbLoI).
While we are of course delighted that our undergraduate degrees are getting more popular, we are also seeing very strong growth in our introductory statistics courses, with currently over 5,000 students taking these courses during an academic year. To deal with this growth and also ensure that our courses provide a solid introduction to statistical practice and reasoning, we are in the middle of implementing a number of changes to our undergraduate programs, including some welcome personnel changes and additions, and the introduction of an online data analysis component.
Picture: Undergraduates representing majors and minors from the department.
New Undergraduate Advisor
We recently hired Mark Dahlke as a new assistant professor. Mark just completed a PhD in Statistics in our department and has 18 years of experience teaching mathematics and statistics. He will become the new undergraduate advisor, taking over the duties of Jennifer Hoeting, and teach several large undergraduate courses.
In accordance with guidelines established by the American Statistical Association for undergraduate education, the majority of our introductory courses now incorporate a data analysis component. In these sections, “Statcrunch” (http://www.statcrunch.com) is the tool used for data analysis. Statcrunch is a simple, intuitive, web-based program that is available at no cost for our students enrolled in these classes. This allows the students to perform real data analyses and experience first-hand how statisticians manipulate data to obtain answers to scientific questions. Rick Gumina, our long-time instructor, was instrumental in getting Statcrunch adopted in these courses.
Getting to Graduate School
Professor Jennifer Hoeting, our long-time undergraduate advisor, runs a "How to get to Graduate School" program for our students each fall. Check out her webpage on the many opportunities for undergraduate students including internships, summer programs, graduate school scholarships and more (www.stat.colostate.edu/~jah/summer/summer.html). We continue to send an impressive proportion of our undergraduates to top graduate schools. In the past several years we have sent students to Iowa State University, University of North Carolina at Chapel Hill, University of Washington, Yale, and more. Many of these students receive full-ride scholarships. Another popular choice among our undergraduates is to continue on at CSU to pursue a graduate in Statistics from our department.
We’d love to hear from you
If you are one of our undergraduate alumni, whether with a degree in Statistics or Mathematics, we’d love to hear from you! Drop us a note at email@example.com to let us know you how you are doing, and tell us how your CSU education shaped your career.
Sun Bak Hospital, a graduate from the statistics department at CSU (M.S. '10), traveled to North Korea with the apolitical organization Statistics Without Borders last summer. She was one of approximately 12 summer faculty members to teach statistics courses at Pyongyang University of Technology and Science (PUST). The summer program, Pyongyang Summer Institute (PSI), is a 4-week program designed to provide statistics courses (primarily survey statistics related courses) to students in the DPRK where resources are severely limited.
She volunteered the first two weeks to teach introductory statistics and statistical computing to undergraduate and graduate students at PUST. During the two weeks of PSI, she stayed at a comfortable guest house on campus (PUST is a boarding school), had meals with the students at the school cafeteria and participated in off-school activities such as playing volleyball. The students at PUST are very bright, well disciplined and very eager to learn. The students have no access to the outside world, and only graduate students have restricted access to the internet - reserved exclusively for peer-reviewed journal searches in a highly monitored environment. Authorities strictly monitor lecture materials, lectures, and topics of conversations. Despite these restrictions and cultural differences, she enjoyed the two weeks in North Korea. It afforded an opportunity to participate in soft diplomacy through science by sharing knowledge and experience as a statistician and in software development. For Sunny, it was a most rewarding experience on many levels. She looks forward to participating in the PSI program next year. To learn more about the PSI program and read the stories of other faculty members of PSI, visit http://www.pyongyangsummerinstitute.org/
Sunny lives in Honolulu, Hawaii working as a statistical consultant and a software developer.
Picture: Sunny Bak (far right) with other Statistics Without Borders affiliates.
Honglin Hu (on the far right in the picture) is a transfer student from East China Normal University and a senior majoring in Mathematics with a Concentration in Statistics at CSU. She was selected to attend an undergraduate workshop at the Statistical and Applied Mathematical Sciences Institute (SAMSI) in May.
The week-long program was held at SAMSI in Research Triangle Park, and at North Carolina State University in Raleigh, North Carolina. The workshop was an introduction to the field of uncertainty quantification (UQ) to students from the mathematical and statistical sciences. The overall goal of the workshop was to illustrate the need for and power of quantitative methods to confront the very hardest and most important data- and model-driven scientific challenges.
The students worked on controlling specific stochastic dynamical systems, specifically, shock absorber systems for cars on bumpy roads. They acquired a working knowledge of techniques such as Kalman filters and Monte Carlo method. The students worked in groups of three under the supervision of SAMSI researchers and postdocs.
In Honglin’s own words: “I heard about the Undergraduate Workshop at SAMSI from Professor Cooley, and he encouraged me to apply. During the workshop, I met a lot of people, from peers and postdocs to professors, who shared similar interests as me. We learned about the techniques of Kalman Filter and Monte Carlo, and we used Matlab to help us model how to make a car ride smoothly on a bumpy road. I enjoyed the process of mastering the theory part first and then applying the methods. It also reminded me of the great importance of mathematical knowledge like Linear Algebra and ODE in engineering and statistical applications. I am very grateful to have been able to participate in such a fabulous workshop for undergraduates andI will continue pursuing my passion in statistics!”
Picture: (left) Honglin Hu (far right) with other students at the workshop. (right) Dan Cooley, Professor of Statistics, encouraged Honglin to apply.
Department Welcomes Two New Faculty Members
We are delighted to welcome Gerwyn Green and Darren Homrighausen in the Department of Statistics.
Gerwyn graduated with a Bachelor's degree in Mathematics from the University of Wales, Swansea. He obtained a Master's degree in Statistics from the University of Missouri, Columbia and a PhD at Lancaster University under the supervision of Peter J. Diggle. His PhD focused upon multiple testing and the development of novel linear mixed models for the analysis of gene expression data. His statistical interests are in mixed modeling including longitudinal data analysis and joint modeling of longitudinal and survival data. In his spare time, Gerwyn adores being outdoors. He is a cycling enthusiast, both road and mountain, a keen fly fisherman and a technically inept skier. (Those are his own words!)
Darren has a Bachelors degree from the University of Colorado in Economics and Mathematics, and a MS and PhD in Statistics from Carnegie Mellon University, where he worked under the supervision of Christopher Genovese. His main applied research interests are in the application of statistics to problems in astronomy and cosmology, working on projects related to astronomical imaging, supernova classification, and detection. Outside of statistics, Darren enjoys hiking, running, weight training, playing and watching sports, and coffee.
Gerwyn and Darren were both hired in newly created positions, made possible by the addition of the MAS degree [add link to other story in newsletter] to the graduate offerings of the Department. Gerwyn will be in charge of managing the MAS degree and will also work as a statistical consultant in the Graybill Statistical Laboratory. Darren is in a regular tenure-track position, not directly affiliated with the MAS program, and will provide much needed additional research and teaching capacity to the Department.
Each year, the Department honors its best students in a number of categories with awards made possible by the generosity of our alumni and friends. The winners of these awards are selected by the faculty in the Department. Seven students shared five awards this year:
Remmenga Scholarship in Applied Statistics: Grant Weller
Boes Award for Excellence in Teaching: Wade Herndon
Boardman Statistical Consulting Award: Stacy Edmondson
Graybill Award for Excellence in Linear Models: Ela Sienkiewicz and Lulu Wang
Madison Memorial Award: Ephraim Hanks and Yuan Wang
Congratulations to our award winners!
Picture: Award winners from 2012. Left to right: Ela Sienkiewicz, Wade Herndon, Stacy Edmondson and Lulu Wang
Distance Education in Statistics Evolves in New Direction
We have been offering the MS in Statistics by distance since 1992, and over 100 students have completed their degree with us since then. After 20 years and beginning in Fall 2012, we are phasing out the MS and replace it with the MAS as the Department’s distance offering. This is being done because we believe that the shorter, more focused courses of the MAS degree and the more hands-on training provided in those courses correspond more closely to what most of our distance students are looking for in their graduate education. So far, the initial interest in the new degree as been nothing short of overwhelming, and we have accepted almost 40 new students into the initial offering of the program.
Technological Changes Improve Delivery of Distance Courses
Additional changes to our distance learning program involve improvements to our distance classrooms and use of technology. We now have access to two classrooms with state of the art equipment. In fact, these are the first two classrooms in the nation with this type of technology. The classrooms each have two Smartboards installed for instructor use. The instructor writes their notes on the primary Smartboard and then ‘moves’ the notes to the secondary Smartboard after they have filled the screen. A podium at the front of the room contains a computer for the instructor to use with the Smartboards, as well as a Sympodium. The instructor is able to write on either the Sympodium or the Smartboard. They may also run data analyses or access internet sites from the computer and display these on the primary Smartboard screen for students to view. The primary Smartboard screen is captured to video during the video taping process. New recording technology (Echo360) allows us to show the Smartboard capture in a large portion of the viewing window, with a smaller picture of the instructor showing, as well. This allows for a clearer display than what we’ve been able to achieve in the past with our older recording technology, hence allowing the students to take more accurate notes. So far student reaction to this new technology has been quite positive.