"Designing for Truth, Scale and Sustainability" - WSSSPE2 Keynote

kaitlin thaney@kaythaney ; @mozillascience

WSSSPE / 16 nov 2014

designing for truth, scale and sustainability

doing good is part of our code

help researchers use the power of the open web to change science’s future.

(0)

power, performance, scale

our current systems are designed to create

friction.despite original intentions.

current state of science

articlesdata

patents

some have a firehose

articlesdata

patents

traditions last not because they are excellent, but because influential people are averse to change and because of the sheer burdens of

transition to a better state ...

“

“Cass Sunstein

downside of output-driven recognition systems

“There’s greater reward, and more temptation to

bend the rules.”- David Resnik, bioethicist

let’s look at an example

2004-2010$350 million spent

$60m+ for “management”70+ tools created

(1)

the culture of the specialist is changing.lowering barriers to entry.

- access to content, data, code, materials.- emergence of “web-native” tools.- rewards for openness, interop, collaboration, sharing.- push for ROI, reuse, recomputability, transparency.

“web-enabled research”

what do we mean by “open research”?

community technology practices

collaborative interoperable open review

participatory discoverable data management

recognition open tools sharing / reuse

mentorship designed for reuse

documentation / versioning

we’re facing a perception crisis.

at the sacrifice of scientific progress.

Source: Wolkovich et al. GCB 2012.

“... up to 70 percent of research from academic labs cannot be reproduced, representing an enormous waste of money and effort.” - Elizabeth Iorns, Science Exchange

http://xkcd.com/1445/

instill best (digital,

reproducible) practice

“research hygiene”

(2)

shifting practice takes a multi-faceted approach.

designing for scale.

research social capital capacity

infrastructure layers for efficient, reproducible research

open toolsstandards

best practicesresearch objectsscientific software

repositories

incentivesrecognition / P&Tinterdisciplinarity

collaborationcommunity dialogue

trainingmentorship

professional devnew policiesrecognition

stakeholders: universities, researchers, tool dev, funders, publishers ...

our systems need to talk to one another.

focus on environments that foster reuse, progress.

code as a research objectwhat’s needed to reuse ?

http://bit.ly/mozfiggit

(community driven)metadata for software discovery: JSON-LD

http://bit.ly/mozfiggit

http://softwarediscoveryindex.org/report/

Instead of cancer driving the development of technology, it was the development of technology that drove caBIG moving into position where this technology could be

adopted by individuals who were interested in cancer.

- Andrea CalifanoColumbia University

“

“

[Their] approach to fulfilling [their] mission was upside down.

- Andrea CalifanoColumbia University

““

(3)

our practices are limiting us.

how to further adoption of open, web-enabled science?

“web-enabled science”- access to content, data, code, materials.- emergence of “web-native” tools.- rewards for openness, interop, collaboration, sharing.- push for ROI, reuse, recomputability, transparency.

“web-enabled science”what’s missing?

- access to content, data, code, materials.- emergence of “web-native” tools.- rewards for openness, interop, collaboration, sharing.- push for ROI, reuse, recomputability, transparency.

current activity:235+ instructors

(60+, training)4000+ learners

rethinking “professional development”

lowering (not exacerbating) barriers to entry

fostering a (sustainable) community of practitioners

(4)

we need to create safe spaces for our software.

designing with sustainability in mind.

“nested sustainability”

KTucker, CC-BY-SA-3.0

We are used to billion-dollar software, and it’s not what we can afford.

I am worried that unless we rein in our expectations, we will do this experiment again and we will get the same result ...

- Joe GrayOregon Health + Science University Center

“

“

1. design with the audience, not individual,

in mind.(be careful of the snowflake syndrome.)

2. design to unlock latent potential of our systems.

(and keep in mind engagement.)

3. rethink how we reward researchers and support

roles.

4. be mindful of jargon/semantics traps.

we’re here to help.teach, contribute, learn.

http://[email protected]

[email protected]@kaythaney ; @mozillascience

special thanks: