1. Overview.

Research to create and improve statistical and quantitative methodologies is central to all of science. The scholars who do this research tend to be scattered across many disciplines in any university, as is certainly the case at UC Merced. Many universities have statistics or biostatistics departments as a focus for such work, but UC Merced is not likely to develop such a department for many years. Consequently, creating a common identity for this kind of research at UC Merced requires another venue.

Eighteen months ago, an interdisciplinary group of UC Merced faculty began to organize the Center for Statistical and Quantitative Research (CeQR) to meet this need. Its development was impeded during the 2012-13 academic year when a key organizer went on sabbatical for that entire year. CeQR’s development is now back on track, with work already underway to establish a CeQR website and to develop both undergraduate and graduate minor areas in statistics.

A Center for Statistical and Quantitative Research would provide a venue for the development of interdisciplinary research projects, the sharing of expertise, as well as valuable graduate training.  Ideally the Center would have its own statistics lab, which would provide resources, including potentially training materials and a meeting space, to facilitate innovative quantitative research and interdisciplinary collaborations.  A speaker series could bring in scholars with expertise currently lacking on campus, and thus could greatly enhance the quantitative training available to both faculty and graduate students.  The participation of faculty from across disciplines in seminars and workshops would foster the exchange of knowledge across disciplines.  UC Merced does not currently have a comparable venue for these sorts of exchanges. Also, the Central Valley does not have a comparable quantitative research center, yet would definitely benefit from its presence.  UC Merced faculty have received numerous contacts from Valley community members seeking assistance with research, especially survey research.  A quantitative research center would thus not only benefit UC Merced faculty, but would also potentially serve a serious need in the community.

CEQR will implement a number of activities over the years as it becomes established: (1) an interdisciplinary quantitative colloquium speaker series, (2) an undergraduate interdisciplinary minor in statistics, (3) improvements to the graduate statistics curricula that might eventually result in a quantitative minor area at the graduate level, (4) a website that will list CEQR activities, a list of pertinent faculty with their quantitative interests, and a list of pertinent quantitative courses at both the graduate and undergraduate level, (5) interdisciplinary extramural research and training grants in statistics and quantitative research, (6) collaborative interdisciplinary quantitative research, (7) a possible summer workshop series that might serve as a source of income for CEQR and a source of training for UCM and other graduate students, (8) eventual infrastructure improvements such as advanced statistical/quantitative computers and software in a dedicated lab, and a statistical consultation clinic where graduate students could both be supported financially and offer consulting services to others, (9) funds for students at all levels to attend quantitative/statistical workshops around the country, and (10) eventual funds for faculty and graduate students to submit grant proposals for seed money for research.

2. Faculty and Expertise

CeQR faculty come from all three schools. Below, we list each faculty by school and area, along with a brief summary of their quantitative interests.

School of Engineering

Electrical Engineering and Computer Science (CeQR Faculty: Miguel Carreira-Perpinan, Ming-Hsuan Yang, Shawn Newsam) The Electrical Engineering and Computer Science (EECS) program has a concentration of faculty who perform research on quantitative data analysis, particularly on large, complex data also known more recently as Big Data. Professor Newsam develops pattern recognition techniques with a current focus on satellite imagery and computational biology simulations. Professor Carriera-Perpinan focuses on machine learning, in particular developing dimensionality reduction algorithms to help discover structure in the complex data. And, Professor Yang performs research in computer vision which has application to a large range of problems which could benefit from automated image understanding. This focus on data analysis naturally lends itself to interdisciplinary collaboration and thus the EECS program would benefit from the cross-disciplinary interactions that a quantitative research center would facilitate. Since the so-called Big Data challenge is only likely to grow with our increased ability to sense our world, the EECS program expects to continue hiring faculty in areas related to quantitative data analysis.

School of Natural Sciences

Applied Mathematics (CeQR Faculty: Harish Bhat, Arnold Kim, Roummel Marcia, Juan Meza (Dean)). The applied mathematics faculty have three strategic academic foci: (1) modeling complex systems, (2) numerical analysis and scientific computing, and (3) data science. In all three of these foci, probability and statistics play a large role. However, it is the third focus in data science that will interface most closely with CeQR. Applied mathematics faculty are currently engaged in data science research problems in image processing, genomic science, and finance among other applications. The applied mathematics faculty is committed to growing the data science focus in its undergraduate and graduate programs. Recently, applied mathematics faculty have been awarded a highly competitive NSF grant to establish the DESCARTES Scholars Program aimed at training top-notch undergraduate students in state-of-the-art training in computational and data-enabled science. As part of this grant, the applied mathematics faculty will introduce undergraduate and graduate courses in probability, stochastic processes, and computational statistics to train students in data science. Applied mathematics faculty are also establishing collaborations with scientists at several national labs to provide additional modes for training students and summer research opportunities in data science. The CeQR will be valuable to the applied mathematics research program because it will facilitate innovative collaboration opportunities, and provide an important means to interface with other researchers on campus across a diversity of research areas.

Quantitative and Systems Biology (CeQR Faculty: Michael Colvin, David Ardell, Emilia Huerta-Sanchez, Emily Moran) QSB faculty use a variety of modern statistical techniques including Baqyesian statistics, nonparametric methods and machine learning to analyze and integrate diverse biological data.

Earth Science (CeQR Faculty: Steve Hart). Ecologists use statistical modeling methods extensively in their work. For example, Dr. Hart does computer simulation modeling regarding the controls of biogeochemical processes and productivity in managed and wildland terrestrial ecosystems. Plans are to add a computational ecologist to the faculty during the current hiring season.

School of Social Science, Humanities and Arts

Cognitive and Information Science (CeQR Faculty: Ramesh Balasubramaniam, Chris Kello, Michael Spivey). CIS faculty have interests in computation and data science. These include computation tools for robotics, modeling the emergence of language, developing neural processing units, computational systems neuroscience, computational linguistics, and to simulate animal foraging patterns.

Economics (CeQR Faculty: Rob Innes, Alex Whalley). Economics is perhaps the most quantitative of the social sciences. Though the economics faculty are not currently involved in developing new statistics, they use a wide array of cutting-edge regression analyses, and are interested in supporting all efforts to increase the quantitative skills available at UC Merced.

Political Science (CeQR Faculty: Courtenay Conrad, David Fortunato, Tom Hansford, Steve Nicholson, Emily Ritter, and Alex Theodoridis). Unlike the vast majority of political science programs, the political science group at UC Merced is entirely quantitative in orientation.  The newly approved PhD program in Political Science is similarly distinctive in its dedication to training students to conduct quantitative research in the areas of political institutions & political economy and political cognition & behavior.  As noted in the separate political science strategic focusing proposal, UC Merced's political science program compares very favorably with top-20 programs in terms of publishing high quality research.  Faculty and graduate students within this group have interests in, for example, maximum likelihood estimation, limited dependent variable models, causal identification, game theory, computational linguistics, "big data," and item-response measurement models.

Management (CeQR Faculty: Fanis Tsoulouhas). Management and Entrepreneurship will emphasize quantitative methods (including modeling, optimization, numerical data analysis, simulation techniques and statistical estimation) in their research.

Psychological Sciences (CeQR Faculty: Sara DePaoli, Jack Vevea, Will Shadish). The quantitative psychology faculty specialize in Bayesian statistics, a cutting edge area in modern quantitative work. No other quantitative psychology program in the nation has that focus. Bayesian statistics is currently a desired area of research for federal funding agencies (e.g., IES), which continues to put us in a good place for securing external funding. Faculty have been quite successful in obtaining extramural funds form the Institute for Education Science and other sources (over $2.1 million since 2005). Because a quantitative psychology specialty is rare in psychology departments, the investment of a moderate number of new lines can result the largest such program in the nation by 2020. Interdisciplinary collaboration is extensive because faculty are asked to consult on projects across many disciplines. Quantitative psychologists need access to high quality computers and programs, and lab space (though less than most other psychologists). By quality and number of faculty, currently comparable peer programs are at UCD, UCLA, University of Virginia, and University of Illinios Urbana Champaign. Aspirational peers are University of Notre Dame, Ohio State, and University of North Carolina Chapel Hill. Because so few quantitative programs exist, graduate student demand exceeds supply of available placements. We currently have nine graduate students over our three faculty, and anticipate increasing enrollment at this ratio going forward.

Sociology (CeQR Faculty: Irenee Beattie, Kyle Dodson, Nella Van Dyke).  The Sociology program includes faculty who have expertise in a range of quantitative techniques.  Faculty members Beattie and Van Dyke both have expertise in event history analysis, also known as survival analysis, as well as advanced knowledge of survey research methodology.  Professor Beattie also uses propensity score modeling.  Sociologist Kyle Dodson has extensive knowledge of longitudinal data analysis as well as cluster analysis.  And Professor Almeida has experience with GIS mapping techniques.  Sociology faculty are engaged in interdisciplinary collaboration on a range of topics.  Professor Van Dyke has worked with faculty in Natural Science (former Dean Pallavicini) and campuswide (Advance Grant project) to provide needed expertise in survey research, and Professor Almeida worked with an Environmental Sciences graduate student on GIS mapping. The Sociology program anticipates adding additional quantitative faculty in the next six years, and a quantitative research center would provide valuable support.  PhD graduates in sociology with advanced statistical skills are in high demand, and thus we anticipate that this training will help our graduates secure high quality positions upon completion of the program.

3. Resource requirements

CeQR will be a new Centralized or Organized Research Unit (CRU or ORU) at UC Merced. No faculty FTEs are needed. Resource requirements are modest, including funding (a) for outside speakers to be part of a colloquium series, (b) a centralized computer facility with high end desktop computers loaded with a full array of software needed by CeQR faculty and graduate students, preferably set up as servers that can be accessed by multiple users simultaneously and remotely, and (c) travel funds for graduate students to attend statistical and quantitative conferences and meeting.

Looking to 2020, UC Merced should consider two additional developments. First might be the development of a research institute at UC Merced, similar to the Health Sciences Research Institute, that would coordinate the many efforts currently underway at UC Merced to advance diverse kinds of statistical and computational work. Such an institute could serve as an umbrella for coordinating and administering existing and proposed centers such as CeQR, including the Center for Human Adaptive Systems and Environments (CHASE), the Center for Computational Biology, and the Center for Theory and Computation. Second, statistical and quantitative research at UC Merced would benefit from the development of a statistical consulting center and a survey research center, and eventually, a statistics department. These may have to be longer range plans given existing resource constraints.

4. Impact metrics

The impact metrics worksheet does not apply well to CeQR because CeQR does not propose a program with FTE and enrollment. Rather, impact for CeQR will be best determined by (a) member productivity in publications, grants, and conference presentations, (b) success in fostering interdisciplinary collaborations, and (c) development of statistical and quantitative curricula ranging from minor areas to possible master’s degree programs in statistics.