class: center, middle, inverse, title-slide # Data Science Education Dilemmas (and Opportunities) ### Joshua Rosenberg ### 2018-10-16 --- # How I came to STEM Education - Science teacher (in Cleveland Co., NC) - Interested in *context* and educational technology (at MSU) - Worked on a science education research project focused on modeling - Used ESM to study how learners engaged in scientific and engineering practices - Became interested in the role of data when engaging in scientific and engineering practices - Assistant Prof. of STEM Education at UT --- class: center, middle # What is the role of data in education? --- # Different names suggest different roles - Educational data science? - Data science in education? - Data science education technology? --- background-image: url(images/dset.png) background-size: 70% --- # Different names suggest different roles - Educational data science? - Data science in education? - Data science education technology? - **Data science education** --- # Focus on data science education - *[Subject matter] education* implies a focus on teaching & learning - Especially important if aim is to work at the K-12 level - Uncommon - Can be empowering to learners - Turns the tables --- class: center, middle # What is data science? --- background-image: url(images/data-science-model.png) background-size: 60% (Conway, 2013) --- background-image: url(images/data-science-hadley.png) background-size: 60% (Wickham & Grolemund, 2018) --- background-image: url(images/data-science-peng.png) background-size: 45% (Peng & Matsui, 2018) --- # The provenance of data science education - Education - *Data modeling* (Lehrer & Romberg, 1996) - *Studies of statistical reasoning* (Wild & Pfannkuch) - *Statistics* - Breiman (2001) - Donoho (2017) - *Data science practice* - Erickson (2017-2018) - McNamara (2016) - *Curriculum reform documents* - NGSS Lead States (2013) and National Governors Association Center for Best Practices, 2010) --- # What's not a dilemma? - Working with data *is* empowering - Young children *can* do data science - Equity *is* a first principle .footnote[[*]see the statistics curriculum in New Zealand] --- class: center, middle # What is a data science education dilemma? --- # Modeling data science education dilemmas - Generated dilemmas - Received feedback from a peer - Grouped into three themes (as a preliminary model for these dilemmas): - Purpose - Pedagogy - Policy --- # Purpose - Unplugged vs. with (computational) tools - Mastering one (computational) tool vs. learning general principles - Use of explanatory vs. predictive models - Creating data vs. using existing data sets --- # Pedagogy - Focus on modeling and visualizing data vs. creating and structuring data - Tools for learning vs. real tools - Taught as part of math and science courses vs. part of statistics or computer science courses --- # Policy - Resource-intensive tools vs. access and expense - Undergraduate and advanced degree vs. K-12 --- class: center, middle # What are some opportunities? --- # Opportunities - Availability of existing curricula - Lots of attention in industry and at post-secondary level - Rare area of overlap between math and science standards - Goal-directed use of engineering and computer science ideas - Demanding, open-ended, potentially meaningful - Some strong equity focus among some practitioners - Not as commercially-oriented as computer science education initiative --- # Some examples to follow - Michelle Wilkerson and colleagues - Victor Lee and colleagues - Pratim Sengupta and colleagues - Hollylynne Lee - Lynn Hodge and Joy Bartlett - Dan Anderson --- # Thank you! .pull-left[ Homepage: [**http://joshuamrosenberg.com**](http://joshuamrosenberg.com) Slides created via the R package [**xaringan**](https://github.com/yihui/xaringan) ] .pull-right[ <img src = "images/jmr-utk-headshot.jpg"> ]