Problem-Focused Immersion (PFIs)
Many data-driven decision-making processes—where numbers and narratives are intertwined—share several attributes including: (i) ambiguous cause-effect relationships; (ii) inherent nonlinearity and feedback; (iii) emergent properties that dominate system behavior; and (iv) imprecise predictability. To be effective in such a decision-making process, Data Professionals must learn how to integrate objective numbers with subjective narratives. The growing calls for such an integration have not come with an instruction manual or a general theory of practice. We propose a problem-focused and data-driven set of PFIs to address this challenge. An overarching theme of these program elements is to show how a chosen problem can be used to advance foundational ideas and theory in data science for improved decision making by focusing on several interrelated activities described below.
Interdisciplinary Immersion Program (PFI-1)
This immersion program is designed as a team-based problem-solving experience with rapid iterations, feedback, and revision to apply and enhance students’ skills and learning in an intense, immersive data-focused experience. This program is inspired and informed by effective educational experiences that bring diverse teams to address difficult challenges. Project-based courses, where the primary goal is to research a problem and create a demonstrable prototype over the course of a semester, are a key educational innovation and provide transformative experiences for students. These project activities are often designed to support a client who seeks solution for a specific problem. These courses provide “sprint” style iteration and refinement of ideas and processes, opportunities to “pitch” ideas, and develop prototypes for testing and implementation. This process occurs with rapid iteration and ambitious development timelines. Key professional skills are developed via these experiences while technical skills are refined and enhanced. However, while such project-based formats are common in Business, Product Design, and Software Development courses, they remain the exception in STEM fields, and are often reserved for “capstone” projects. Future Data Professionals need opportunities like these to develop team-based competence and receive feedback on their holistic performance that supports key skills in both self- and team-awareness, evaluation, and management.
The motivation for this program element comes from successful models in data-driven language learning (University of Maryland), computer science (Johns Hopkins), and science policy translation (University of California, Santa Barbara), programs to identify, develop, and implement interdisciplinary research themes within the context of short and intense engagements. A goal of our proposed immersion program is to show how a chosen problem can be used to illustrate the full “virtuous circle” of data driven decision making. Activities includes hackathons (that have been very successful at Tufts) and interdisciplinary team-based problem-solving workshops. This immersion program provides a forum for team-based exploration of a problem from multiple perspectives, identifying possible solution pathways given the capacity and constraints imposed by the context, and then producing context-sensitive solutions that are actionable with measurable outcomes. Such a forum also provides opportunities for participants to “pitch” their ideas, present preliminary findings and prototype their data-driven tools to an invited panel of professionals who can evaluate the suitability of the proposed solution, judge the appropriateness of the data-driven techniques employed, and rate the clarity of the team’s communication skills.
This Interdisciplinary Immersion Program consists of two colloquia (Fall and Spring) and an intensive week-long workshop. It brings together students and faculty from different disciplines as well as practitioners from our national and international partners. A key goal of this immersion program is to show how a chosen problem (e.g., Flint Water Crisis) can be used to illustrate how interrelated activities—data to discovery → discovery to insights → insights to action—can lead to effective data-driven decision making. Each Interdisciplinary Immersion Program will focus on a prototypical theme. D3M@Tufts Fellows and Certificate seekers are required to participate in at least one immersion program during their tenure at Tufts. The Interdisciplinary Immersion Workshop is envisioned as the culmination of an extended learning process, starting at the beginning of the prior year’s Fall semester. This allows the team to explore initial ideas via semester-long (Fall Colloquium) activity, then plan more specific activities over the next semester (Spring Colloquium) with a chosen problem. This immersion program leverages logistical support from Tufts DISC initiative.
Bi-weekly Data Colloquium (PFI-2)
Colloquia provide an opportunity for fellows to present evolving research ideas to the group. Students receive meaningful and focused feedback that helps clarify their research development. We have found that sharing early-stage research in a low-risk interdisciplinary environment encourages innovation and creates space for the formation of collaborative partnerships. The bi-weekly colloquium also offers an opportunity for all participants to develop a common data science vocabulary (e.g., “vector” has a different meaning in “mosquito as a vector for transmitting malaria” compared to “vector as a mathematical attribute with magnitude and direction”) and shared knowledge for interdisciplinary data driven decision making.
During our first Fall Colloquium, D3M@Tufts faculty, students, and invited speakers from industry will present a set of problems, starting with these guiding questions: (a) Why is this an important data-science question? (b) What is the societal relevance of this question? (c) How do we go from what we know or do not know (based on the available scientific theory and observational evidence) to how we make decisions (actionable policy/implementation options with measurable outcomes)? (d) What data do we have (or not have) to address this problem? (e) What are the contextual constraints for this problem? and (f) Given the current science and the current context (i.e., “what we know now”), what are decisions/actions that are feasible? At the end of the semester, colloquium participants, D3M@Tufts faculty, students, and invited speakers will together choose a problem for the Interdisciplinary Immersion Program. The Spring Colloquium will be designed to develop deeper understanding of the chosen problem with an extensive discussion of existing literature and consultation with data professionals from the industry as well as from our national and international partners.
A Data-Focused Independent Capstone Project (PFI-3)
Data-focused thinking is as necessary for professional development for future data professionals as the ability to read and write. Evidence-based thinking and thoughtful evaluation of varying or conflicting interpretations of data are necessary ingredients to distill insights from the deluge of data. Data-focused thinking skills are the critical thinking skills required to filter the signal from the noise and evaluate the relationship between data, its interpretations, and definition of “fact.” Students seeking a D3M@Tufts Certificate are required to work on an independent data-focused project. This project allows students to examine and evaluate data and data products relevant to their discipline as well as expand and reinforce cross-contextual data interpretation skills from other disciplines. This data-focused project should fit into a student’s research and count as an MCE elective credit for this project’s completion. Our goal is to assist and promote high quality, degree-relevant HDR experiences and project outcomes within the timeline and program requirements of each student across diverse graduate programs and departments.
This culminating data-driven independent student project requires students to make use of acquired MCE competencies within their disciplinary context. The project provides D3M@Tufts Certificate seekers an integrative understanding of data by synthesizing numbers and narratives as well as quantitative and qualitative perspectives. This project emphasizes integrative data focused thinking and data-driven decision making. Each student may incorporate appropriate data-derived products to better appreciate strengths, weaknesses, and evaluation of different data-supported arguments ranging from models to meaning and numbers to narratives. Data-focused project activities provides a balance between exploration (development of new solutions) and exploitation (taking advantage of what is known to work) for the chosen datasets. This project experience actively engages students in exploring, analyzing, and interpreting data, with significant latitude to invent, assess, and evaluate conflicting interpretations of data, and to propose actionable outcomes and data-driven strategies for measuring their societal impact.
Annual Data Symposium (PFI-4)
The Annual Data Symposium provides an opportunity for students to present their problem-focused data-driven research. This symposium provides an opportunity to showcase the findings from the Interdisciplinary Immersion Program and Data-Focused Project to the university, Boston community, and a global audience via social/web involvement.