Investigating Data Like a Data Scientist: Key Practices and Processes

Abstract

With a call for schools to infuse data across the curriculum, many are creating curricula and examining students’ thinking in data-intensive problems. As the discipline of statistics education broadens to data science education, there is a need to examine how practices in data science can inform work in K-12. We synthesize literature about statistics investigation processes, data science as a field and practices of data scientists. Further, we provide results from an ethnographic and interview study of the work of data scientists. Together, these inform a new framework to support data investigation processes. We explicate the practices and dispositions needed and offer a glimpse of how the framework can be used to move the discipline of data science education forward.

Primary students’ learning about citizenship through data science

“…The cycle includes (1) problem elicitation and formulation, (2) getting the data, (3) exploring the data, (4) analysing the data, (5) communicating the results, and cycling back to considering the problem at the beginning. Lee and her colleagues ( 2022 ) developed a similar framework that further emphasised visualising and modelling the data (present but not as visible in the IDSSP cycle). These frameworks can provide an accessible vision of what may be adapted to primary and secondary mathematics education.…”

“…Fry and Makar ( 2021 ), for example, drew on the IDSSP cycle to illustrate how key ideas in data science education can be introduced in age-appropriate ways using the current curriculum in mathematics and digital technologies. Further examples of embedding data science in school mathematics have been illustrated in the literature from primary (e.g., Fielding & Makar, 2022 ; Fry et al, 2023 ) to secondary (Lee et al, 2022 ). We recognise that these examples come from well-resourced classrooms with adequate access to technology and experienced teachers.…”

A secret agent? K-12 data science learning through the lens of agency

“…As we suggested in the previous section, data science, too, has grown out of a need to manage new forms of variable data created by rapid growth in computing power (Cao, 2017;V. Lee & Wilkerson, 2018;H. Lee et al, 2022).…”

Visualizing statistical edutainment: What you see is what you get

“…As the upper left cartoon in Figure 1 (captioned by college student John Montagu) states, data visualization is an art. It is an art to find or create a visualization that best tells the story of a study's results, even if it must be combined with the science of statistics to give an appropriate impression, and visualization is "a really underrepresented part of the data science art" [36]. Sometimes novel and creative visualizations profoundly impact society, such as British nursing pioneer and statistician Florence Nightingale's (1820-1910) use of a rose diagram, circular histogram and polar area display to communicate, summarize, and identify causes of death and then led to nursing reform (see song by [41]).…”