Slide 1

Slide 1 text

Useful Tools for Teaching and Outreach in Data Science Stephanie Hicks Assistant Professor, Biostatistics Johns Hopkins Bloomberg School of Public Health Faculty Member, Johns Hopkins Data Science Lab @stephaniehicks

Slide 2

Slide 2 text

We define Data Science Workflows broadly, as: "The reproducible and transparent ways you do your data analysis" This encompasses: • literate code documents (e.g. Rmarkdown, Notebooks, etc.) • analysis scripts • pipelines • environment management tools • etc. Mike Love Tiffany Timbers

Slide 3

Slide 3 text

"The reproducible and transparent ways you do your data analysis" This encompasses: • literate code documents (e.g. Rmarkdown, Notebooks, etc.) • analysis scripts • pipelines • environment management tools • etc. We define Data Science Workflows broadly, as: I’m going to focus on this as it relates to Teaching and Outreach

Slide 4

Slide 4 text

What is a case study? https://ia600203.us.archive.org/11/items/cu31924018826713/cu31924018826713.pdf C. C. Langdell Dean of Harvard Law School from 1870 to 1895

Slide 5

Slide 5 text

What is a case study? Before Langdell's tenure, the study of law was very technical And students were simply told what the law is. During Langdell’s tenure, he applied the principles of pragmatism to the teaching of law à students were compelled to use their own reasoning powers to understand how the law might apply in a given case.

Slide 6

Slide 6 text

What is a case study? “Law, considered as a science, consists of certain principles or doctrines. To have such a mastery of these … is what constitutes a true lawyer; and hence to acquire that mastery should be the business of every earnest student of law.” – C. C. Langdell C. C. Langdell Dean of Harvard Law School from 1870 to 1895

Slide 7

Slide 7 text

What is a case study?

Slide 8

Slide 8 text

What is a case study?

Slide 9

Slide 9 text

What is a case study?

Slide 10

Slide 10 text

What is a case study?

Slide 11

Slide 11 text

Elements of a case study Background: provides context for the problem to be solved Problem: a dilemma to be resolved or a decision to be made Supporting information: data, exhibits, interviews, supporting documentation, etc

Slide 12

Slide 12 text

Characteristics of a good case study Real real-world situations / real problems / based on real events / realistic, complex, and contextually rich situations / contemporary / recent / tells a real story Focused on students engages students / student centered / students make choices / active learning Link between theory and practice application of concepts in practice / bridges the gap between theory and practice / make choices about what theory to apply / highlight connections between academic topics and real-world situations / connects the academy and the workplace Ambiguous complex and ambiguous / present unresolved issues, situations, or questions / ”art of managing uncertainty” / without a detailed script / coping with ambiguities

Slide 13

Slide 13 text

Teaching

Slide 14

Slide 14 text

Guidelines for Assessment and Instruction in Statistics Education (GAISE) College Report 2016 1. Teach statistical thinking. (Teach statistics as an investigative process of problem-solving and decision making). 2. Focus on conceptual understanding. 3. Integrate real data with a context and purpose. 4. Foster active learning. 5. Use technology to explore concepts and analyze data. 6. Use assessments to improve and evaluate student learning.

Slide 15

Slide 15 text

Case studies in data science? The American Statistician, Vol. 53, No. 4, pp 370-376 “The model calls for … substantial exercise[s] with nontrivial solutions that leave room for different analyses.

Slide 16

Slide 16 text

Case studies in data science? The American Statistician, Vol. 53, No. 4, pp 370-376 Elements of a “case study”: • Introduction • Data • Background • Investigations • Theory

Slide 17

Slide 17 text

https://opencasestudies.github.io

Slide 18

Slide 18 text

https://jhu-advdatasci.github.io/2018/ http://cs109.github.io/2014/ http://datasciencelabs.github.io/2016/ (Harvard University – CS – over 400 students online, in person – 25 TAs – Python) (Harvard SPH – Biostats -- 150 students – online, in person – 8 TAs – R) (Johns Hopkins SPH – 25 students – in person – PhD Biostats – 2 TAs – R)

Slide 19

Slide 19 text

OCS: elements of a data science case study •Motivation • What is the question? What is the context/background for the question? •What is the data? •Data import •Data wrangling •Exploratory data analysis / data visualization •Data analysis •Summary of results

Slide 20

Slide 20 text

Demo https://opencasestudies.github.io

Slide 21

Slide 21 text

Incorporating Git/GitHub workflows into data science workflows in the classroom

Slide 22

Slide 22 text

http://happygitwithr.com

Slide 23

Slide 23 text

Describes how to integrate Git/GitHub workflows into statistics courses targeted towards students with computational backgrounds

Slide 24

Slide 24 text

Jacob Fiksel @jfisel1 https://arxiv.org/abs/1811.02021 Describes how to integrate Git/GitHub workflows into statistics courses targeted towards students with non-mathematical backgrounds (e.g. public health, sciences) “Using a reproducible workflow in statistics is vital to a complete data analysis, yet for faculty and students with limited computing background, learning version control tools such as Git can be difficult and intimidating. [Here] we outline some of the ways that the Git workflow can be implemented in statistics courses at all levels”

Slide 25

Slide 25 text

SDSS 2019 CS01 - Teaching Statistics More Effectively to a New Generation of Students Thursday 10:30am-12:05pm Using GitHub with Statistics Undergraduates Jo Hardin, Pomona College h/t @AmeliaMN

Slide 26

Slide 26 text

Incorporating Slack into the classroom

Slide 27

Slide 27 text

No content

Slide 28

Slide 28 text

Slack in the Classroom

Slide 29

Slide 29 text

Using Slack for Communication and Collaboration in the Classroom Albert Y. Kim, Smith College bit.ly/slack_sdss

Slide 30

Slide 30 text

Leonardo Collado Torres

Slide 31

Slide 31 text

Slack for Conference Program Committees!

Slide 32

Slide 32 text

Outreach

Slide 33

Slide 33 text

No content

Slide 34

Slide 34 text

No content

Slide 35

Slide 35 text

No content

Slide 36

Slide 36 text

No content

Slide 37

Slide 37 text

No content

Slide 38

Slide 38 text

No content

Slide 39

Slide 39 text

No content

Slide 40

Slide 40 text

No content

Slide 41

Slide 41 text

Acknowledgements Rafael Irizarry Mike Love Tiffany Timbers Leah Jager Margaret Taub Leonardo Collado Torres

Slide 42

Slide 42 text

Feel free to send comments/questions: Twitter: @stephaniehicks Email: [email protected] #rladies Thank you! Normal distribution Weibull distribution Poisson distribution