This is an archive copy of a page that previously appeared at:​

https://wiki.80000hours.org/index.php/Data_science​
Last update: December 2015

Data science

This page contains all notes on data science as a career path over and above what we put in our career profile.
Read the profile first, here.

Contents [hide]

​
​Profile type
​What is this career path?
​What are the people like?
​Role impact
​Direct impact potential
​Earnings potential
​Advocacy potential
​Career capital
​Common exits
​Culture
​Exploration value
​Personal fit
​Entry requirements.
​What does it take to progress?
​Barriers
​Job satisfaction
​Alternatives
​Academia
​Past experience
​Take action
​Learn more
​Next steps
​Self learning resources
​Bootcamps
​Best resources
​Remaining issues
​Research process
​
​
This is an archive copy of a page that previously appeared at:​
https://wiki.80000hours.org/index.php/Data_science​
Last update: December 2015
​

Profile type
Exploratory

What is this career path?

Introduction to what data science is

Data scientists analyse data (often very large datasets - see big data) using advanced techniques drawn from
the disciplines of computer science and statistics in order to deliver actionable insights for organisations. Data
science jobs are related to data analyst jobs -- the main difference between these is that data analysts don't
use advanced techniques from computer science and statistics in order to perform their analyses. Data
scientists tend to hold PhDs in quantitative subjects.

Drew Conway's Venn Diagram of data science

For most problems you can use existing algorithms - 90-95% of the time.

One of our users said "a data scientist is a statistician who knows a big more software engineering than normal;
or a software engineer who knows a bit more statistics than normal."

What are the people like?

This is an archive copy of a page that previously appeared at:​
https://wiki.80000hours.org/index.php/Data_science​
Last update: December 2015

This varies widely by company, city and team. We've been told that at highly skilled teams in San Francisco the
people are highly talented, hold PhD's in quantitative from elite universities and the environment closely
resembles academia, with very cerebral discussions.

Role impact

Direct impact potential

Earnings potential
http://www.oreilly.com/data/free/files/2014-data-science-salary-survey.pdf

http://www.payscale.com/research/US/Job=Data_Scientist_%2f_Engineer/Salary (compared to software

engineer: http://www.payscale.com/research/US/Job=Software_Engineer/Salary)

Glassdoor - Average Base Salary: $104,476, number of job openings 3,449 (Software engineer - $98,074,
Number of Job Openings: 104,828) (Mobile developer - $79,810, Number of job openings $79,810)

http://www.glassdoor.com/blog/jobs-america/

https://datajobs.com/big-data-salary

"There are in fact edge cases of data scientists getting paid over $250,000 in unique situations –
e.g. hedge funds, or special cases of advanced algorithm development – but this well above the
norm."[1]

Advocacy potential

Career capital

Common exits
The two main routes for progression within data science are:

●​ Management roles - data scientists can become valuable product managers because they know
the limits of data science and software engineering. Well-rounded individuals can become senior
leaders.
●​ Individual contributor roles - these are when you are an expert in a specific area
Some fields of academia are easier to go back into after working in industry than others. Computer Science is
easier to re-enter. Astrophysics is very difficult to re-enter.
This is an archive copy of a page that previously appeared at:​
https://wiki.80000hours.org/index.php/Data_science​
Last update: December 2015

Culture

Exploration value

Personal fit
We've been told that the most important skill for a data scientist is the ability to think like a scientist and is the
slowest and hardest to train. Second most important is ability in probability and statistics, and the third is
programming.

"If you don't love data for its own sake, then you will find it hard to compete with such candidates. Burtch,
however, says everyone should learn to love data, if only for the sake of their career. "Within 10 years, if you're
not a data geek, you can forget about being in the C-suite," Burtch says." [2]

Entry requirements.

What does it take to progress?

The two most important qualities for progression suggested to us in an interview with a senior data scientist
were:

1.​ Being an excellent communicator

2.​ Strategic ability - having insight into what will most efficiently get an organisation to meet its
high-level goals

Barriers

Job satisfaction
The only downside of the job, Greenberg says, is the time spent "cleaning" data — pruning it to remove
irrelevant findings. "That part's not that exciting and you spend a lot of time doing it," he says.
(http://mashable.com/2014/12/25/data-scientist/)

Vacation policies tend to be good because of how in demand the skillset is.

Alternatives

Academia
This is an archive copy of a page that previously appeared at:​
https://wiki.80000hours.org/index.php/Data_science​
Last update: December 2015

We've been told that in data science what you work on has direct impact on something tangible. Some
companies will have data scientists working at the edge of human knowledge and making original contributions,
though at some companies and roles data scientists just apply current domain knowledge to achieve
organisational goals.

Past experience

Take action

Learn more

Next steps
For getting a job the most important skills to gain are really strong knowledge of statistics and probability, a
statistical programming language like R or Python, and the basics of SQL. R is more for high-level research,
whereas Python has broader uses and is a larger investment as there is more to learn.

Making a portfolio of projects is important for demonstrating your ability. You should start doing your own
projects and write them up on a blog to show off your visualisation and communication skills, and put your code
in a GitHub repository.

Self learning resources

http://datasciencemasters.org/

https://github.com/zipfian/data-science-primer

https://www.udacity.com/course/data-analyst-nanodegree--nd002

https://www.coursera.org/specialization/jhudatascience/1?utm_medium=courseDescripTop

https://www.mysliderule.com/learning-paths/data-analysis

Bootcamps

Lists of bootcamps:

http://datascience.community/bootcamps

http://yet-another-data-blog.blogspot.co.uk/2014/04/data-science-bootcamp-landscape-full.html

Should you do a bootcamp? The placement statistics are impressive. One of our members said Insight was the
"the best educational experience of my life".
This is an archive copy of a page that previously appeared at:​
https://wiki.80000hours.org/index.php/Data_science​
Last update: December 2015

Best resources
Linear Digressions - Udacity's podcast on machine learning and data science

Remaining issues

Research process
Sources of info

http://en.wikipedia.org/wiki/Data_science

BLS

http://www.bls.gov/careeroutlook/2013/fall/art01.pdf

http://www.bls.gov/ooh/computer-and-information-technology/computer-and-information-research-scientists.htm
#tab-1

Udacity

Skills checklist for getting your first data analyst job (pdf)

4 types of data science jobs

Hiring Trends in Data and What They Mean for You

Quora

http://www.quora.com/What-is-the-Data-Science-topic-FAQ

http://www.quora.com/How-do-I-become-a-data-scientist

Journal articles

http://www.nature.com/naturejobs/science/articles/10.1038/nj7546-253a

Media

So you wanna be a data scientist? A guide to 2015's hottest profession (Mashable)

A Very Short History Of Data Science (Forbes)

Data Scientist: The Sexiest Job of the 21st Century (HBR)