Information for Academic Statisticians in Mathematics Departments and for Reviewers of Academic Statisticians in Mathematics Departments

People who write letters for tenure or promotion of academic statisticians may find it useful to cite portions of the Mathematical Association of America (MAA) Guidelines and the American Statistical Association (ASA) Response to these guidelines. Included below are portions of these documents that may be most useful. Candidates who stand for review may also find this information helpful.

Also included below are suggestions for both candidates and reviewers based on discussions at an Isolated Statisticians meeting and at a Statistics in the Liberal Arts Workshop meeting.

From the MAA Guidelines for Programs and Departments in Undergraduate Mathematical Sciences (http://www.maa.org/guidelines/guidelines.html) and the ASA’s Response to MAA’s guidelines (http://www.maa.org/guidelines/asa%5Fresponse.html):

“Some consulting and other professional activities may advance the scholarship and teaching of faculty members and the department. Consulting and other professional activities may fit into the category of teaching or scholarship and in that case should be evaluated accordingly, or such activities might be evaluated as a separate category, with correspondingly less emphasis on other categories. Supervision processes and evaluation procedures for formal consulting activities should include the monitoring of faculty progress in maintaining and improving the quality of these activities.” (MAA C.8.e)

“The best way to encourage active student-faculty interactions and to enable faculty to give students individual attention is to provide a small-class environment with fewer than thirty students in each section. Also with restricted class size, faculty members gain flexibility to adopt a teaching style that best fits both the material to be learned and their students' needs. (MAA D.2.b)

“Departments should encourage and assist faculty members who investigate, try out, and evaluate alternative teaching techniques that show promise in helping some students be more successful in learning in the mathematical sciences.” (MAA D.5)

“When a newly hired faculty member with a graduate degree in statistics is assigned to teach a mathematics course, such as calculus, special consideration should be given for preparation time.” (ASA Response to MAA C.2.d)

“For the applied statistician, statistical consulting with scholars in other disciplines constitutes an extremely important area of research since these experiences are a primary way in which the statistician keeps abreast of the latest in statistical methodology as well as keeping active in the skills of data analysis and statistical reasoning.” (ASA Response to MAA C.3.d)

“One further point needs to be made concerning the obligation for communication of results. Statisticians communicate the results of their scholarly activity to other members of the research team. Not infrequently these results indicate that the data do not support the desired hypotheses or that the experiment was poorly designed and the data cannot be used to evaluate specified hypotheses. In many such instances, the results of the scholarly activity of the statistician cause the experimenter to refrain from improperly submitting research for publication. These instances are of great service to the scholarly community. In these cases, timely communication of results to peers (the experimenter) cannot be measured in number of publications. The fact that the publication did not occur should be recognized as a positive peer evaluation by the experimenter of the statistician’s contribution to the project.” (ASA Response to MAA C.3.d)

Notes/Suggestions made at Meetings of Isolated Statisticians

• Curriculum development and educational research can constitute creative/scholarly work.  Such work can result in peer-reviewed journal articles (e.g., in JSE or SERJ) or lead to other forms of peer review (e.g., on CAUSEweb).

• Breadth in research is actually good and expected of statisticians (i.e., a statistician’s research may not be narrowly focused).

• Statistical consulting raises some important issues.

• When starting a major consulting project it takes time for a statistician to learn the new field(s).

• Statistical consulting can be mistakenly seen as service when in some cases it’s actually scholarship.

• Collaboration (e.g., a multi-authored paper) is a valuable activity for a statistician.

• To get credit for consulting (especially for unpublished work in industry and work with proprietary data), a candidate should keep a detailed record of the work and solicit supportive letters from people with whom the candidate has collaborated.

• It is important for a candidate to have expectations for consulting and criteria for evaluation placed in writing by the dean and/or the department (see MAA.C.5.d, MAA C.8.e, MAA.C.8.f, ASA’s response to MAA C.2.b).

• Candidates and reviewers may want to read and cite the article “Statistical Consulting is Scholarship” by William Wilson in The American Statistician (1992, vol. 46, no. 4, pp. 295-298).

• Candidates (and reviewers of candidates) at liberal arts colleges may want to read and cite the article “Statistics at Liberal Arts Colleges” by Tom Moore and Rosemary Roberts in The American Statistician (1989, vol. 43, no. 2, pp. 80-85).

• In evaluating an academic statistician, the reviewer need not be an expert in the statistician’s research area (especially when writing to people who aren’t specialists—like a tenure committee of a liberal arts college). For example, the reviewer can assess the quality of the journals of publication, the type of methodology (e.g., new statistical results versus interesting applications to new areas), and the impact on other research (e.g., number of citations of the work).