1 of 44

Building �Data Science�in K-12 Education�State of the Field

Data Science 4 Everyone

The University of Chicago

Summer 2024

2 of 44

All Companies

All Top Jobs

Recap on the “Data Revolution”:

All Industries

Source: Future of Jobs Survey 2020, World Economic Forum.

https://www3.weforum.org/docs/WEF_Future_of_Jobs_2020.pdf

3 of 44

Does anyone how much data there is?

4 of 44

175+ Zettabytes

2 Zettabytes

The answer is growing exponentially:

5 of 44

= 1 Megabyte (MB)

6 of 44

23.6 billion miles

Grows 2x by 2030.

Our digital footprint has outpaced physical.

7 of 44

More data on data:

5.4 Billion Internet users connected globally.

5 Billion videos viewed on Youtube per day.

30 Trillion+ estimated web pages.

180 Billion emails sent per day.

Source: VentureBeat. https://venturebeat.com/business/how-google-searches-30-trillion-web-pages-100-billion-times-a-month/; Internet Live Stats: http://www.internetlivestats.com/

8 of 44

Today’s workers use data regularly:

Source: Survey distributed via Freakonomics tweets and web site; respondents not representative—skewed male, higher income, higher education

9 of 44

And many wish they had more in school:

Source: Survey distributed via Freakonomics tweets and web site; respondents not representative—skewed male, higher income, higher education

10 of 44

New landscape analysis:

1 in 4 jobs require basics of data science

Sources: 2019-2022 LinkedIn, GlassDoor, and DOL SCOC Skill Codes

11 of 44

New landscape analysis:

1 in 2 jobs require some “data analysis”

Sources: 2019-2022 LinkedIn, GlassDoor, and DOL SCOC Skill Codes

12 of 44

13 of 44

New landscape analysis: sectors

14 of 44

A reminder:

138 college disciplines

now require data and

statistics coursework

Sources: American Statistical Association

15 of 44

Source: https://www.buzzfeed.com/carolynkylstra/hype-in-health-journalism

Source: https://www.theguardian.com/science/2018/apr/12/one-extra-glass-of-wine-will-shorten-your-life-by-30-minutes

Separating fact and fiction

is now very… difficult.

16 of 44

Automated Prediction Model

Prompt (text data in)

“Tell me if wine is good for me or not…”

Answer (text data out)

“Three cups of wine per day is recommended for a sharp mind, fit body, a sound spirit…”

17 of 44

Automated Prediction Model

Prompt (text data in)

“Tell me if wine is good for me or not…”

Answer (text data out)

“Three cups of wine per day is recommended for a sharp mind, fit body, a sound spirit…”

(training data at large n)

(validation)

(general model for repeated queries)

Data Science Learning => AI Readiness

18 of 44

19 of 44

20 of 44

“Resveratrol has been studied for its potential therapeutic use,^[13] with little evidence of anti-disease effects or health benefits in humans.”

Resveratrol is a type of polyphenol.

Wikipedia says “Polyphenols were once considered as antioxidants, but this concept is obsolete.”

WedMD says “polyphenol-rich diets… have lower death risks.”

I’ll walk through what I noticed, and what I had some questions about:

The Wikipedia citation is from a single study in 2004.

WedMD seems trustworthy, it’s written by medical Doctors… but then a little more digging reveals that there have been historical allegations that there articles are consistently biased towards recommending health supplements and other pharamaceuticals when little or very minimal evidence exists to support those conclusions.

https://www.vox.com/2016/4/5/11358268/webmd-accuracy-trustworthy

So, how would you go about answering this complicated question? You can ask your doctor… and of course if you’re trained in data science, you know the importance of large samples, of analysis that doesn’t rely on outlier results, of repeated analysis over time on different types of people… to help inform your diagnosis, if you really want to get to the bottom of this. Another person, without those basics, would throw their hands up, and do what they wanted to do anyways... Whether that was to down a Merlot every night, or tell their husband their going to have heart disease if they drink a glass with dinner.

This problem scales to many others - whether it’s important health decisions, planning large financial purchases, or deciding how to vote and what information to trust.

21 of 44

Critical Mission:

Catalyze the adoption of data science & data literacy as an integrated component

of K-12 education

by 2030.

22 of 44

23 of 44

My path in personal life:

Youngstown, OH

Silicon Valley

24 of 44

My path in personal life:

25 of 44

My path in personal life:

What is the imagery here? ��What made these differences? Economics. And supply / demand. Youngstown use to be wealthy b/c of its location.

A major steel manufacturer, ½ way between Chicago and New York, with iron ore in the nearby Appalachians.

When the steel-mills shut-down, our community was devastated b/c the workforce was not prepared to switch quickly. Everyone knew how to make steel. They were not educated in anything else that was adaptable. ��As we enter the information age, now the wage premium is on digital. Computing, data, e-commerce, cybersecurity, quantum, blockchain, etc.

The technology tool or buzz-term of the day may change, but I don’t think the idea of “digital information” or ”data” is going away anytime soon. It’s quite fundamental. ��So let’s run with this metaphor a bit more. “Data is the new oil” was coined back in 2006. Controversy on who said it first but it stuck. It’s still true today.

26 of 44

The old map of crude oil:

U.S. Geological Society / American Geosciences Institute, Historical Oil & Gas Production, 1859-2005

27 of 44

The new map: “data is the new oil”

28 of 44

“Data is the new oil”

Primary determinant of wealth?

In 20^th century: physical resources -> $$$

In 21^st century: knowledge resources -> $$$

29 of 44

We need to help students�prep for the opportunities of �a data-fueled world:�

Artificial Intelligence

Blockchain / Web 3.0

Quantum Computing

Internet of Things (IoT)

Machine Learning

Cyber Threat Detection

ChatGPT

E-Commerce & Trade

(HTF) High-Frequency Trading

AV (vehicles, manufacturing)

30 of 44

What if we asked students directly?

31 of 44

32 of 44

What if we asked students directly?

33 of 44

The need is clear. How are we addressing it?

34 of 44

11 Focus Groups

120+ edu. leaders

�33 States, D.C., and Canada

�K-5, 6-8, 9-12, 2-year, 4-year

�Urban, Suburban, Rural

�Public, Private (including Religious)

Building a National Data Learning Progression

35 of 44

Integrating in Core School Subjects:

36 of 44

Data Science & Literacy Act of 2023

37 of 44

Data science is growing:

38 of 44

Data science is growing:

39 of 44

NCSSM is a national leader �– you’re in the right place:

40 of 44

Building modern data education:

Elementary (K-5)

Middle School

(6-8)

High School

(9-12)

College

(2-/4-Year)

			Data Science Degrees	Data Science Careers
		Data Science Techniques		Modern Careers
	Data Literacy (all subjects)			21^st Century Life
	Algebra & Probability (math / stats)			Building Block
Data Intuition				Building Block

Real-World

(Life)

41 of 44

With relevant tech tools:

Elementary (K-5)

Middle School

(6-8)

High School

(9-12)

College

(2-/4-Year)

			Multiple Software	Data Science Goals
		Intro to Data Analytics Software		Data-Related Careers
	Spreadsheets / Everyday Software			21^st Century Skills & Jobs
Education-Specific Software				Building Block

Real-World

(Life)

42 of 44

43 of 44

44 of 44

44