ASEE NCS Conference 2019

Full Program »
PDF File
View File
pdf
2.4MB

Bringing Computational Science to Collegiate Education

Data science has been heralded as the “sexiest job of the twenty-first century.” IBM estimates that the number job openings for US data scientists will increase from 364,000 to 2,720,000 by 2020. Motivated by increased demand for computational skills in science and industry, universities have started offering programs in data science, but are faced with the challenge of developing courses in hard-to-define areas such as “computational modeling” and “data analytics.” Loosely constructed, data science is the science of collecting and analyzing data and is closely related to computational science, which is interested in applications of computing. Sometimes data science is cast as an enhanced version of statistics or computer science, but the predominate academic view envisions computational and data sciences as a unified, yet multidisciplinary field. Practices such as data analysis, mathematical modeling and programming are common in mathematics, statistics and computer science, but computational science adds applications to these practices to foster a particular worldview about modern uses of computing.

Defining “data science” and “computational thinking” provides a framework for talking about computational science in the context of science and mathematics education. New academic programs at the collegiate level represent different approaches to infuse computational science into the curricular landscape. The high cost of creating new programs leads many universities to create programs using resources from existing departments, often computer science and statistics, while other universities have invested in entirety new departments and coursework. Unfortunately a lack of research about computational science education makes it difficult to establish curriculum or choose teaching methods, although progress has been made defining terms, developing broad standards, and reviewing practices used in industry. Evidence-based teaching practices in STEM provide a starting point for new collegiate programs in computational science interested in promoting effective teaching. In particular, the literature strongly supports using active learning, problem-based learning and peer instruction to improve student engagement. However, these methods may be difficult to transfer directly from computer science and mathematics because of cultural and pedagogical differences, necessitating new research on computational science education.

Mitchell Eithun
Michigan State University
United States

 



Powered by OpenConf®
Copyright©2002-2018 Zakon Group LLC