The University of Michigan’s Consulting for Statistics, Computing and Analytics Research, established in 1946, provides free software support and consulting on data analysis for all University researchers in need of assistance. This month, the University disclosed they will phase out the unit.
According to an August 13 newsletter from the Office of the Vice President for Research, OVPR leaders framed the closure as a restructuring to better serve the entire campus.
“This transition is necessary to help OVPR serve a wide range of needs on campus more effectively,” the newsletter read. “This transition will occur in a phased manner to minimize the disruption to those obtaining statistical consulting from, and those working for, the unit. For investigators who have contracted for statistical consulting with the unit, and investigators who currently have dedicated effort on a sponsored project through the unit, OVPR will arrange continued support.”
According to data provided to The Michigan Daily by CSCAR Interim Associate Director Hyungjin Myra Kim, the unit fielded more than 2,300 consulting requests between January and November 2024. While researchers in the Medical School accounted for the largest share at 17%, faculty and students across LSA, Public Health, Engineering, Information and other units also make heavy use of this service.
Kim said in an interview with The Daily CSCAR was not consulted by Arthur Lupia, vice president for research and innovation, about the decision. Concerns from community members prompted a petition signed by graduate students and postdoctoral students urging the administration to keep CSCAR open.
“At no point were we informed of any desired changes or areas of concern,” Kim said. “It would have been valuable to have the opportunity to introduce him to our mission, highlight our contributions and share what we value as a team.”
In an interview with The Daily, Brady West, former CSCAR employee and research professor in the survey research center, echoed the same frustration. He said the plan for closure came as a surprise, especially since according to Kim’s data, the unit’s entire operation costs the University the equivalent of three and a half full-time staff members.
“I know for a fact that CSCAR was blindsided by the decision,” West said. “I have no idea why OVPR did not reach out to former employees to discuss this further and I still have not heard any justification for the decision. CSCAR’s budget is literally a drop in the bucket relative to the vast amount of service that it provides.”
In an email to The Daily, Lupia defended the decision, saying the changing campus research landscape justified the closure.
“The decision helps to better align the office’s focus on emerging research requirements and ensure resources are directed to areas where the research office has unique responsibilities,” Lupia wrote. “In past decades, CSCAR was the only campus source for statistical consulting. That is no longer the case, and we are coordinating with campus partners and others to offer these alternatives going forward. In the coming weeks, we will roll out new arrangements for obtaining these services.”
Community members said there are few comparable alternatives. The Michigan Institute for Clinical and Health Research provides similar data support, but charges for services at a base cost of $93 per hour for biostatistical support.
Kentaro Toyama, WK Kellogg Professor of Community Information and Senate Advisory Committee on University Affairs member, compared CSCAR’s role in statistics to another well-established campus support center.
“CSCAR is like the Sweetland Writing Center for statistics,” Toyama said. “CSCAR provides in-depth methodological guidance about using statistics for research to the whole university community. Nothing (is) comparable, as far as I know, either in accessibility or quality.”
West warned that the closure could undermine researchers’ ability to compete for funding and publish their work.
“The ability to write CSCAR into research grant proposals, receive free advice on research methodology or seek assistance when revising the statistical aspects of peer-reviewed journal articles that have received revise and resubmit decisions will all disappear,” West said. “Frankly, the entire research enterprise across the three campuses will suffer significantly, and the quality of the statistical work that (the University) is doing will decline.”
In an interview with The Daily, Rackham student Sidney Xiang said she has found CSCAR invaluable in her research work.
“Many of our research advisors refer us to CSCAR on a regular basis for statistical analysis issues or just to get a second opinion on stuff,” Xiang said. “A lot of data issues that come up in research are more subtle and idiosyncratic than what we encounter in coursework, and it’s really useful to see how the CSCAR staff approach problems in addition to the recommendations they give at the end.”
Daily News Editor Emma Spring can be reached at sprinemm@umich.edu.
