Faculty FTEs in statistics at UW-Madison have dropped even as demand for statistics instruction (Figure 1) and research collaboration has dramatically risen. We must employ short-term lecturers to teach nearly all introductory statistics courses and some advanced undergraduate courses. At the same time, the TA budget has barely kept up with undergraduate demand, and has been severely cut for graduate courses. In order to provide quality instruction in data analysis using modern tools, we need additional faculty to translate new research into the classroom, and supporting staff to provide infrastructure to leverage the limited faculty we have. Over the near to long term (5-10 years), we propose hiring 3 FTE of staff and 5 FTE of faculty to address these needs.
The Statistics Department teaching load in the past five years has been 72 lectures per year, based on 18 faculty FTE, but the load recently rose by 4 to 76 due to increased demand in introductory (301, 371) and intermediate (311, 312) courses. While this growth and demand is great, reflecting the interest and appreciation in statistical methods for experimental design and data analysis, this increased demand has an impact on the Department. Most of this instruction (~40 lectures) is for undergraduate instruction, and most of those (>25) are taught by short-term lecturers. Our current faculty FTE is 14.75, which, at 3 courses per FTE covers only 44 lectures. The addition of 5 new faculty FTE would greatly increase faculty teaching of undergraduates while allowing some modest increase in new course offerings at both the undergraduate and graduate levels.
The target of 19.75 faculty FTE is more modest than our 2002 vision, which had a goal of 21 FTE, reflecting the challenging economic landscape. Note that we could have up to 5 FTE of retirements in this same time frame; new faculty proposed here would be in addition to replacement of retiring faculty. We are also making calculations on the presumption that in the near future, a 1 FTE faculty teaching load will be three courses per year. This change is necessary to remain competitive in terms of recruiting and retaining faculty. [See successful 2011 Teaching Load Reduction Request for documentation of peer institutions.]
New statistics faculty would likely include joint faculty. Statistics traditionally aims to hire the best candidate without regard to specialization, as we value the ability of faculty to adapt to changing research demands. As we gradually increase our faculty, we will place especial emphasis on big data research methods and on research that involves complicated data structures, which conforms to broader changes in the field of statistics. This new emphasis on skills in researching data structures touches on traditional realms of other data science programs. Therefore, one or more new faculty hires would involve substantial appointments in Computer Science, Industrial & Systems Engineering, or other data sciences programs, or in interdisciplinary ventures including the Wisconsin Institute of Discovery and the Morgridge Institute of Research. This represents a revival of George Box's 1960 vision of a theoretical hub with distributed statistics across campus. The concept has evolved over the intervening 50 years, and contributed to the success of this Statistics Department. However, at this juncture, we need a re-injection of resources to remain vital. Other top statistics programs across the country are moving in this direction (e.g. University of California, Berkeley, Stanford University, and University of Chicago). The University of Wisconsin-Madison Department of Statistics can surpass them if we are clever.
The Statistics Department will leverage its limited faculty by hiring new academic staff with expertise in instruction, research and advising. The goal is to leverage creative activity, particularly in development of data analytic tools, by speeding up translation of new methods into practice in collaborative research and in research-like instructional venues that impact our students as well as students, staff and faculty across campus. Specific immediate areas are detailed below (PVL job descriptions being drafted). Other innovations will emerge over time as we develop new support mechanisms for faculty and staff to do their jobs more effectively. Here is a brief summary of the proposed positions:
- Undergraduate advising and oversight of undergraduate instruction (1 FTE): teach statistics courses at introductory or intermediate levels; oversee and redesign Tutorial Service, including supervision of assigned TAs; oversee introductory statistics pedagogy, including supervision of Head TAs; advise undergraduates about the statistics major, including liaising with other campus advising organizations and SOAR; build community among undergraduates interested in statistics.
- Instructional computing and tools for data analysis (1 FTE): teach statistics courses at introductory or intermediate levels; develop, coordinate and teach modules for new Stat 327 “Learning a Statistical Language”; develop data analysis pedagogy to assist TAs in introductory statistics courses with use of statistical computing tools; develop web-based resources for students, staff and faculty regarding statistical computing and data analysis tools.
- Research computing and big data analytics infrastructure (1 FTE): develop and teach intermediate and advanced modules of Stat 327; develop research computing infrastructure to support big data analytics beyond the terabyte range; assist faculty and staff to bring research tools into undergraduate and graduate research experiences. Address increasing needs to administer and manage big data and complicated data structures, with attention to data (both collected data and created software) provenance, maintenance and access/retrieval, and to web tools for team collaboration and documentation.