Your Textbook Can’t Help You Here: Applying traditional methods to atypical data
Student projects are a common method of concluding classes or programs through a descriptive analysis and write-up of material. In most statistics courses, the technique or model for completing these projects is known, given, or researched. While answers vary, a student’s output follows a guideline that fits within the framework of a textbook. Further, when a professor assigns a project, he or she has a general idea of what to expect from the analysis.
Such familiarity is not the theme behind the required master’s project course at the University of Massachusetts (UMass) Amherst, which instead involves the learning, executing, and presenting of techniques across all types of inference. Each semester, faculty members solicit current or ongoing research projects from their peers across campus. Pairs of graduate students are matched with research projects and the associated professor based on mutual interest in the material. Each cohort also includes a member of the statistics faculty who oversees the development and execution of the data analysis, with weekly class meetings fostering an atmosphere of collaboration and integration. For details, see the sidebar “UMass Stat 691.”Some content is only viewable by Chance Subscribers