Jessica Utts, Workgroup Chair

Workgroup Members

Jessica Utts (chair), University of California, Irvine, ASA Board

Lori Thombs, University of Missouri, ASA Section on Statistical Education

Dalene Stangl, Duke University, ASA Caucus of Academic Representatives

Morteja Marzjarani, Saginaw Valley State University, SPAIG Committee

Nandini Kannan, The University of Texas at San Antonio, ASA Committee on Outreach Education

Devan Mehrotra, Merck Research Labs, ASA Career Development Committee

Eileen King, Cincinnati Children’s Hospital Medical Center, ASA Advisory Committee on Continuing Education

Marlene Egger, University of Utah, ASA Committee on Applied Statisticians

Keith Crank and Rebecca Nichols, ASA staff representatives

Are statisticians at all levels receiving the education they need to fulfill the roles expected of them in the 21st century? The ASA’s 2010 Education Workgroup was appointed by 2010 ASA President Sastry Pantula and asked to “facilitate a significant discussion among academic units, industry statisticians, and government statisticians about the preparation of statisticians.” The goal was to recommend a process by which the undergraduate statistics curriculum guidelines approved by the board in 2000 could be updated and a document to provide curriculum guidance for MS and PhD degrees in statistics could be created.

The initiative was motivated, in part, by a discussion with the Caucus of Academic Representatives, in which members recognized the need for sharing ideas and information about statistics degree programs across institutions. Recognizing the need for input from a broad spectrum, Pantula appointed representatives to the workgroup from a range of ASA committees.

Members of the workgroup sought information from academic, industry, and government statisticians, mostly through events held at JSM 2010. They found there was substantial interest in the development of the proposed guidelines in the form of desired learning outcomes, rather than a list of recommended courses. Individual departments with degree programs could then decide how best to align their curriculums with those learning outcomes, taking local resources, interests, and, possibly, local employment opportunities into account.

In addition to being used for developing and revising degree programs, the recommendations could be used by external reviewers as benchmarks for program reviews, by departments and programs to justify resource requests to their administration, and by programs that require the assessment of learning outcomes as part of accreditation reviews.

The needs expressed by government and industry representatives went beyond what statistics and math departments currently offer. More than once, they indicated the technical (“hard”) skills will get you an interview, but the “soft” skills will get you the job (and allow you to keep it). Many times, members of the workgroup heard that learning outcomes for statistics degree programs should include the ability to communicate effectively to nontechnical audiences, work as part of a team, integrate information from a variety of sources, and provide technical and nontechnical advice about a project from start to finish. At JSM 2010, Pantula and SAS CEO Jim Goodnight emphasized the need for a strong foundation in core, communication, and computational skills.

Workgroup members learned that the most diverse offerings among training programs in statistics are at the undergraduate level. That also may be the level for which curriculum guidelines are most useful, with the master’s level next in importance. An option gaining in popularity and needing separate consideration is the professional science master’s degree in statistics. This degree is popular among working professionals and often offered partially or fully online. It tends to have less training in mathematical statistics than a traditional master’s degree and is usually a terminal degree. As more universities develop this degree, it would be helpful to have input from potential employers about what knowledge and skills are most useful to include.

PhD programs that train future faculty members tend to be the most focused, presumably because faculty are more familiar with how to train their own future colleagues than with how to train statisticians for industry and government employment. Few faculty members have experience in these sectors, especially at the level of jobs requiring a bachelor’s or master’s degree. Yet, many students who graduate with a PhD in statistics do not enter academic jobs, and guidelines produced from discussions with industry and government statisticians would benefit PhD programs as well.

The process of developing guidelines for programs at all levels should include obtaining feedback from government, industry, and academic employers, as well as former students who currently work for those employers. This latter group could provide feedback on what was missing in their training that would have been helpful in their jobs.

The current guidelines for undergraduate programs were developed through a lengthy process, including a three-day, in-person meeting involving approximately 30 participants. Members of the workgroup anticipate a similar lengthy process will be needed for updating these guidelines and developing new ones for other degree levels.

Recommendations were made to the ASA Board for how to proceed, including a multiyear timeline with a focus on different sectors each year. Anyone interested in getting involved in the development process should contact Jessica Utts, workgroup chair, at jutts@uci.edu.