The Value of Compassion in Learning Data Science

University of Pittsburgh Love Data Week

Ted Laderas

2/13/23

Bio

  • Former academic that loves teaching
  • Data Scientist/Bioinformatician 20+ years
  • Not afraid of Clinical Data
  • Bioinformatics Trainer at DNAnexus
  • Psychological Safety for Everyone
  • Ready for R
  • Data Analytics
  • Cloud for HPC Users
  • Gradual Introduction to Shiny
  • OHSU BioData Club
  • Cascadia-R
  • cvdRiskData

TL; DR

  • You all are smart and brilliant people
  • Don’t beat yourself up while learning
  • Find your crew who you can learn with safely
  • When possible, make things public

My Advice to Grow as a Data Scientist

  1. Have Values
  2. Meet people
  3. Learn Things Together
  4. Build Things Together
  5. Take Care of Yourself
  6. Share Things in Public

1. Have Values

Why Values?

  • Values help you say yes
  • Values Help you say no

My Values

  • Curiosity (about learning and others)
  • Compassion (for myself and others)

There is more than one path

  • Question the PI model
  • Collaboration is the most important

Be Compassionate and Gentle With Yourself

  • Data Science is Hard
  • Don’t Replay Negative Tapes
  • Find people to Learn With

Learning Self-Talk from Students

It’s okay not to know how to do something, practice makes perfect!

It will be hard to learn, but so much fun and worth it.

Dedicate couple hours for practice a day

Learning Self-Talk

It’s going to be okay, it’s really overwhelming right now but Ted is patient and thorough.

It’s a good thing that you’re looking at multiple sources of information about R (i.e. a brief Code Academy stint, the KCRB R4DS group, R Bootcamp links), it will just serve to solidify the topics that get covered.

Don’t forget to Google, remember to ask questions.

This is What it Takes

The 5 minute rule for learning

  • If something doesn’t make sense, take a walk or a break and come back to it.

It’s Dangerous to Learn Alone

  • Use Social Learning to learn faster
  • Having a group makes learning easier

2. Meet People

Who Should You Meet?

  • Fellow Learners!
  • Potential Collaborators
  • Potential Employers
  • “Weak Ties”

An Introvert’s Guide to Networking

  • Networking is not about promoting yourself
  • It’s about being curious about other people
  • Finding common ground and seeing if you want to work with them

Be Curious About Other People

  • Practice Informational Interviewing
  • Don’t overthink your questions
  • Be genuinely interested in what people do
  • Don’t limit yourself to your field

Do or Attend Lightning Talks!

  • Lightning talk: 5-10 minute invitation to chat
  • Best way to advertise what you’re interested in
  • Best way to meet people with your same interests

Volunteer at Conferences

  • Low-key way to meet people
  • Working on things together is super helpful
  • Example: TA for Building Tidy Tools Workshop

Building Tidy Tools - Thanks Ian and Emma!

3. Learn Things Together

Compassion = Psychological Safety

Kahn defines Psychological Safety as:

Psychological safety is being able to show and employ one’s self without fear of negative consequences of self-image, status or career.

Lowering Psychological Burdens for Students

Kahn, Psychological Engagement of Personal Engagement and Disengagement from Work, https://journals.aom.org/doi/abs/10.5465/256287

Psychological Safety -> Effective Learning

You need to find it for yourself

Start or Join a Group!

  • Find your like minded peers
  • Help each other learn
  • Find groups that have Codes of Conduct

Why are Codes of Conduct Important?

  • Establishes the tone for the group
  • States unacceptable behavior

BioData Club Code of Conduct

Example: OHSU BioData Club

  • Started as a group that wanted to learn more data science
  • Everyone is Equal (PhD, Postdoc, Faculty)
  • Working through problems together

BDC: Learners as the best teachers

  • You remember how hard it is to learn
  • Coaching/pedagogy

The best teachers

{.fragment width=600}

https://biodata-club.github.io/materials

The best Teachers

{.fragment width=600}

https://biodata-club.github.io/materials

The Best Teachers

{.fragment width=600}

https://biodata-club.github.io/materials

The Best Teachers

{.fragment width=600}

https://biodata-club.github.io/materials

What did we learn?

  • It just takes a desire to learn
  • Coaching/Pedagogy is important
  • Faculty Sponsorship is Vital

https://biodata-club.github.io/materials

Other Learning Opportunities

Tidy Tuesday

  • Weekly Dataset
  • Learn from other’s code
  • Share your visualizations/code
    • #TidyTuesday #RStats

https://github.com/rfordatascience/tidytuesday

R for Data Science Learning Community

https://rfordatasci.com

Data Visualization Society

https://www.datavisualizationsociety.org/

R-Ladies

https://rladies.org

CSVConf

https://csvconf.com/

The Carpentries

https://carpentries.org/

Mozilla Open Leaders

https://foundation.mozilla.org/en/initiatives/mozilla-open-leaders/

4. Build Things Together

Compassion in Teaching

Collaboration is Fun and a Vital Skill

  • Academic Consortia
  • Open Science
  • Industry
    • It is everything

My Rule for Collaboration

If I meet you and we hit it off, we should work on something together.

Cascadia R Conf: Empower Others

Cascadia R 2023

We started with 6 people and got this

R-Bootcamp

https://r-bootcamp.netlify.app

Ready for R

  • Free for all!
  • Leverage Posit Cloud (free)
  • Videos
  • Notebooks

Ready for R Signup Link

Testimony

As a person fairly new to R, and someone who is slowly completing your Ready for R course, I appreciate your explanations. I have some pre-existing knowledge of R, but this is a great way to reinforce my learning. Thank you!

Testimony

Taking this course has explain[ed] SO many things and made so many things clear. Even just understanding organizing data or file paths. It was a lot of things I didn’t have the language or understanding to ask about

Testimony

I am very happy that because of this class I can build up my self-confidence to work in R. It helps me a lot. With the help of all the exercises, I can practice R on my own data too. Thank you so much. Want to learn more :)

Survey Results

Lots of Users All Over the Place!

5. Take Care of Yourself

Managing Up

  • Communicating with your manager what you need
  • Educating them about what they’re curious about

You Can’t Please All 6 of Your Bosses

  • During COVID
  • I put together 5 new classes in 1.5 years for different departments
  • Also on 3 other grants (lots of other work)
  • I needed something new

You are a Limited Resource

  • Use your values to decide whether something is worth focusing on
  • Use your values to explore possibilities
  • Be compassionate for yourself, because no one else will

Self Care and Self Compassion in Academia

Managing Burnout

  • Burnout is a systemic bug
  • Needs to be managed at multiple levels
  • You by yourself can only go so far

Asking for Help to Get Better

Look for Alternatives

  • Asking around
    • Cohort of people in my network
    • Comparing Notes
  • Weak Ties
    • Lead to DNAnexus

DNAnexus Academy

  • Being confused is part of my job
  • Asking Questions
    • Product
    • Customers

UK Biobank RAP community

  • 500K Participant Dataset
  • Genomics on the cloud
  • Plan learning paths

From HPC

To Cloud

Bash for Bioinformatics

https://laderast.github.io/bash_for_bioinformatics/

6. Share Things

The Unreasonable Effectiveness of Public Work

Yes, but…

  • Social Media is a Double-Edged Sword
    • Especially for BIPOCs
    • Your mental health matters
    • Be aware and use blocking tools

Low-key Share Your Work

  • Tidy Tuesday
  • Data Science Portfolio

Slack/Discords

Is Twitter Worth it Anymore?

  • Yes, but…
  • With Caveats (blocking)

Coming Back to the Beginning

My Advice to Grow as a Data Scientist

  1. Have Values
  2. Meet people
  3. Learn Things Together
  4. Build Things Together
  5. Take Care of Yourself
  6. Share Things in Public

Expressing my Gratitude

Thank You (ggplot version)

Gratitude Code

  • Thanks to everyone for inspiring me and/or your chats!
library(tidyverse)
gratitude <- readr::read_csv("gratitude.csv") |> arrange(First)
col_num <- rep(1:6, each=21)[1:nrow(gratitude)]
row_num <- rep(seq(1:21), 6)[1:nrow(gratitude)]
gratitude <- gratitude |> mutate(name=paste(First, Last), col_num = col_num, row_num=-row_num) 

ggplot(gratitude) + aes(y=row_num,x=col_num, label = as.character(name), check_overlap=TRUE, hjust=0) + geom_text() + theme_minimal() + xlim(1,7) +
  theme(axis.text = element_blank(), axis.title = element_blank(), panel.background = element_blank())

Questions?