Panel presentation at the 2015 Digital Library Federation (DLF) Forum with 5 CLIR Postdoctoral Fellows working in different university libraries to coordinate RDM strategies across campus, Vancouver, Canada, Oct 2015
DLF Panel on RDM Strategies in the Library, Oct 2015
1. TRAINING UP AND
REACHING OUT:
LIBRARY STRATEGIES TO
COORDINATE RESEARCH DATA
MANAGEMENT ON CAMPUS
October 27, 2015
Morgan Daniels
Vanderbilt University
morgan.g.daniels@vanderbilt.edu
@morgand
AnaVan Gulick
Carnegie Mellon University
anavangulick@cmu.edu
@anavangulick
Scout Calvert
UCLA
scout@library.ucla.edu
@windloochie
Sarah Pickle
Assessment Librarian
The Claremont Colleges Library
sarah_pickle@cuc.claremont.edu
@sarahepickle
Stephanie Simms
Research Data Specialist
California Digital Library
stephanie.simms@ucop.edu
@stephrsimms
CLIR/DLF Postdoctoral Fellows in Data Curation
10. IT TAKES A CAMPUS
AnaVan Gulick
Carnegie Mellon University
anavangulick@cmu.edu
@anavangulick
Collaboration across campus to
support research data
Listen to your stakeholders, get
their buy-in, build up your
services accordingly
Teamwork and triage in the
library
14. 14|
2014 Survey Results
2014 Survey, Van Tuyl & Michalek
There is a LOT
of data
Most projects
produce small
amounts of data,
but a few produce
very large amounts
of data
Data storage
and back-up
could be
improved
27. What we did
• 14 staff members, including librarians and
library staff.
• Semi-structured interviews
• Walked through cycle stages to elicit specific
skills, knowledge, and abilities
• Interviews transcribed and coded
32. Thanks to my fellow monks at UCLA:
Rikke Sarah Ogawa, MLIS, AHIP
Bethany Myers, MSLIS
Vessela Ensberg, Ph.D., former CLIR Fellow.
Tony Aponte, MLIS
Apologies to:
Schrader, A. (2010). Responding to Pfiesteria piscicida
(the fish killer): Phantomatic ontologies,
indeterminacy, and responsibility in toxic
microbiology. Social Studies of Science 40(2), 275-306.
Questions, comments: scout@library.ucla.edu
36. 36|
Motivations
Challenge 1: Providing support for
RDM wasn’t in anyone’s job
description.
How to bring them on board and
make them feel this is worth their
effort?
Challenge 2: Not everyone was
ready or confident enough to jump
right in.
How to develop a safe community
that would provide education,
training, and a sense of shared
purpose?
I was conducting outreach with
researchers and support staff to
uncover what help was needed and
what help already existed.
But I had the potential to become a
victim of my own success. I also
needed the support of my subject-
specialist colleagues.
Goal: Enlist my colleagues—who are
often already in researchers’
workflows—to work with me.
39. 39|
RDWG
Planning efforts:
Recruited co-organizers.
Drafted proposal, which included phased-out plan to learn
together, work toward creation of RDM handbook,
implementation of coherent outreach strategy, and ongoing
professional development.
Charged by AD.
40. 40|
YES! But how?
The goals outlined in the RDWG
proposal were well and good, but
they were ours…
How could they best match the
needs of the environment we were
working in?
NB: Question language differs slightly from text of questionnaire.
41. 41|
YES! But how?
Listen.
The goals outlined in the RDWG
proposal were well and good, but
they were ours…
How could they best match the
needs of the environment we were
working in?
NB: Question language differs slightly from text of questionnaire.
42. 42|
YES! But how?
Listen.
The goals outlined in the RDWG
proposal were well and good, but
they were ours…
How could they best match the
needs of the environment we were
working in?
NB: Question language represented here differs slightly from text of questionnaire.
Howoftendoyoutalkwith
facultyandstudentsabout
theirresearchdata?
(n=73)
Howconfidentdo/wouldyou
feelwhentalkingwithfaculty
andstudentsabouttheir
researchdata?
(n=69)
Whatdoyouneedtofeel
moreconfidenttalkingwith
facultyandstudentsabout
theirresearchdata?
(n=37)
14%
23%
38%
25% often
sometimes
rarely
never
43. 43|
YES! But how?
Listen.
The goals outlined in the RDWG
proposal were well and good, but
they were ours…
How could they best match the
needs of the environment we were
working in?
NB: Question language represented here differs slightly from text of questionnaire.
Howoftendoyoutalkwith
facultyandstudentsabout
theirresearchdata?
(n=73)
Howconfidentdo/wouldyou
feelwhentalkingwithfaculty
andstudentsabouttheir
researchdata?
(n=69)
Whatdoyouneedtofeel
moreconfidenttalkingwith
facultyandstudentsabout
theirresearchdata?
(n=37)
14%
23%
38%
25% often
sometimes
rarely
never
9%
25%
28%
25%
14%
very
confident
neutral
not very
not at all
44. 44|
YES! But how?
Listen.
The goals outlined in the RDWG
proposal were well and good, but
they were ours…
How could they best match the
needs of the environment we were
working in?
NB: Question language represented here differs slightly from text of questionnaire.
*Categories created from open-text responses.
Howoftendoyoutalkwith
facultyandstudentsabout
theirresearchdata?
(n=73)
Howconfidentdo/wouldyou
feelwhentalkingwithfaculty
andstudentsabouttheir
researchdata?
(n=69)
Whatdoyouneedtofeel
moreconfidenttalkingwith
facultyandstudentsabout
theirresearchdata?
(n=37)
14%
23%
38%
25% often
sometimes
rarely
never
9%
25%
28%
25%
14%
very
confident
neutral
not very
not at all
51%
22%
help with resources* training/group
45. 45|
RDWG
Mobilize!
May-July 2015
Half a dozen hour-long meetings
held with documentation posted on
group wiki. 10-12 faculty and staff
participants from across PSUL, both
subject and functional specialists.
Topics covered: defining data,
characterizing quantitative and
qualitative data, research data
lifecycle, key aspects of RDM
planning.
RDWG member questionnaire
asked about specific data-
related topics and was
intended to inform future
growth and track progress.
46. 46|
RDWG
Challenges
Hesitant to suggest topics and
volunteer to present.
How can they know what they
don’t know enough to know
that they want to know it?
Fear of scope creep in the
Libraries and “marketing”
services PSUL doesn’t (or
shouldn’t…) provide.
Confirmation of one original
motivation for forming the
group—no clear path for RDM
support in PSUL—while also
resisting an attempt to forge
that path from the grassroots.
47. 47|
RDWG
Going forward
The group is on a temporary hiatus,
partly due to the start of the new
semester, partly due to our hope for
RDM-related commitments the new
ADs, partly due to my leaving Penn
State.
But there’s momentum! Group
members now have a colleagues they
can turn to for support and a sense
of where to find help when they need
it.
Even if the future of RDM training
and community at PSUL takes a
different form, RDWG has broadened
awareness of RDM concerns and
needs on campus while confirming
interest in addressing them.
Maybe it has served as a proof of
concept for future grassroots
organizing in the Libraries.
Sharedvocabulary,
senseofprojectlifecycle
Resourcesongroupwiki
Newsciencedata
librarian
OnboardingtwoADs
interestedinRDM
Uncertain future for RDWG,
but hopeful for RDM support
in the Libraries.
48. 48|
Preliminary
conclusions
I don’t know whether I succeeded—
it’s too early to tell—but the
attendance numbers and diversity
were promising. Group members felt
comfortable asking questions and
acknowledging where the limits of
their understanding were.
The hope is that the information I
gathered can be used not just as
justification for RDWG’s existence,
but also as a way to demonstrate
progress and build confidence in my
colleagues to provide RDM support.
Turningneeds-assessmentinwardhelpedtotakethe
temperatureofmycolleaguesandcreateagroup
thattheywouldvalue.
51. 51|
User Experience/UX
What it is
Why it matters
“aperson’sperceptionsand
responsesthat resultfromthe
useoranticipated useofa
product, system,orservice”
-ISO9241-210
Boston traffic sign image from www.tallahasseewebdesign.com
57. 57|
UX = a strategy for
RDM in libraries
Wikimedia Commons | Arenamontanus
58. DISCUSSION
Q&A
Morgan Daniels
Vanderbilt University
morgan.g.daniels@vanderbilt.edu
@morgand
AnaVan Gulick
Carnegie Mellon University
anavangulick@cmu.edu
@anavangulick
Scout Calvert
UCLA
scout@library.ucla.edu
@windloochie
Sarah Pickle
Assessment Librarian
The Claremont Colleges Library
sarah_pickle@cuc.claremont.edu
@sarahepickle
Stephanie Simms
Research Data Specialist
California Digital Library
stephanie.simms@ucop.edu
@stephrsimms
CLIR/DLF Postdoctoral Fellows in Data Curation
Editor's Notes
Good morning, I am the CLIR postdoctoral fellow in data curation at vanderbilt university. My background is in information science, and for several years my research has focused on barriers to data sharing and reuse and the strategies that teams, individuals, and organizations use to overcome them. I wont be talking about that research today, instead I’ll be discussing my work in the Vanderbilt library, where my job is to design and deliver data management services to the university community. I am focusing on my experience communicating with librarians about RDM- research data management- and some of the outcomes we are seeing on campus from that outreach within the library.
VU is a private university in Nashville, undergraduate enrollment 6,800, graduate and professional school enrollment 5,800, library staff 210, faculty 4,100
Rest assured that while librarians are an important audience for outreach about RDM services, they are not the only audience.
Communication with librarians happens in the context of many other connections with campus administration, IT, faculty staff and students…
With all of these audiences, I want to communicate the what, why, and how of RDM, but in ways that are tailored to their concerns.
Librarians are an especially crucial audience, though, because they have relationships with academic departments that cut across the university, they have relevant expertise in preservation, selection, and metadata creation, and they will likely be asked to take a more central role in RDM services as both the services and requests for assistance grow.
So in terms of their role in communicating about RDM, we might think of librarians near the center of a series of tree rings- an expansion of information outward from the library via librarians, through the course of their work with faculty and students, and tangetially, with the research groups in which those individuals work.
Or we might see this as the champagne waterfall model. The idea that knowledge of RDM is transmitted to librarians, who then send it on to members of the campus community with whom they work. But, of course, this is not a guaranteed flow of information. A librarian who feels less confident in his own understanding of RDM might decide not to broach the topic with faculty and students. And I obviously have no control over the configuration of champagne glasses in this metaphor, some might be missing.
To deal with the first concern, librarian fluency with RDM concepts, I’m holding meetings with librarians, either individually or in groups who work in the same subject areas.
I use these meetings to introduce data management concepts, to learn about faculty with potential interest and need for RDM, and to get people talking together about research data (and perhaps the changing scope of librarianship)
outcomes: offers from libraians to co-present to faculty on RDM and other services and introductions to departmental contacts for people with RDM needs
Another tactic I’ve used is offering workshops on RDM topics. I’ve introduced them as part of the scholarly communications workshop series, which is open to everyone in the campus community- and we’ve had a number of people from the local community attend as well. Our regular attendees are faculty, staff (research and otherwise), students and librarians, who are eager to expand their own digital skills.
In RDM topics, I have offered workshops that covered an introduction to data management and preservation, Data cleaning (OpenRefine), Data mashups (sourcing datasets from various locations), Data visualization
Something unexpected has developed from the workshops-- this series’ relatively high visibility has lead to offers to present to other campus groups
In this network diagram, if the red circle is workshop attendees and the blue circle is people who learned about the workshop online, several of those people are responsible for planning continuing education programs across campus and have asked my colleagues and I to participate, giving us opportunities to make new connections with people who may not have a preexisting relationship with librarians.
At this point, 4 different groups have asked for workshops- a medical informatics group, the IRB, the medical center continuing education program, and a postdoc career development group. These connections are leading to new opportunities for collaboration, requests for data management consultations…
And to a new metaphor I’m considering to think about the communication of RDM throughout Vanderbilt. Neural networks are the pathways between areas of the brain
that are engaged when we learn. The more we practice what we learn and engage those pathways, the more defined they become. Now, this may be a dangerous metaphor to suggest in the presence of my co-panelist, Ana, who is a cognitive psychologist, but we can go into the nuances of the metaphor later. For now, I am suggesting that as we build relationships
around the university in the process of growing and providing research data management services, we are strengthening pathways in the university and creating new ones that can be brought into service for strengthening RDM initiatives and for creating new cross-campus collaborations.
Thank you!
Good morning, I’m Ana Van Gulick and I’m a CLIR fellow for data curation at Carnegie Mellon University in Pittsburgh. My graduate training is in psychology and neuroscience and I’ve been working at CMU for just over a year now at building out of research data management services and infrastructure at the university and library level. CMU is a fairly small institution (7K undergrads ad 7K graduates) but has a robust and divers research program.
Today I’m going to tell you a little about how we’re tackling data management as a community
One key to this is keeping at least part of the discussion going across campus, not just within research groups, and not just within the libraries and listening to each others needs
To facilitate this at CMU, a place where colleges especially can be quite isolated, the Research Data Management Steering committee was formed in 2012.
This group meets regularly with stakeholders across the university to discuss RDM initiatives
It includes reps from the Libraries (both RDM and Schol Comm – with the RDM reps really sponsoring most of the RDM projects), Office of the Vice-Provost for Research (specifically with an eye to research integrity and compliance so complying with funding agency requirements for open data and access as well as generally safeguarding research and preventing data loss), Computing Services (with an eye towards supporting the libraries work in a campus wide fashion – navigating technical solutions for data storage, organization)
We also have research stakeholder representatives to gain their perspective – a faculty rep, a rep from GSA and a director from the undergraduate research office
Together we are building a road map for RDM support that researchers need and want on our campus
What are we building? And who is offering these services?
First it’s important to know what the needs are
In 2014 our then data services librarian, steve van Tuyl undertook a survey of the faculty about their data practices.
I wont go into the results too much here except to say here found there was indeed a LARGE amount of data being produced on campus
However, most projects produce a small amount of data, only 1-10GB. Only the largest 5% of project account for most of the data
Furthermore that solutions for data storage and back up could be improved
He also asked Faculty to prioritize data services they might want to use
He found two main themes – assistance with DMPs – both for grants as well as general research protocols
And Assistance with Data preservation, deposit and documentation
To support our new education and outreach efforts we’ve tried to simplify our communication and center our services around the phases of the data lifecylce
Here is our newly designed data lifecycle which you will find on our website soon, which will be clickable and which pulls out resources and services for each stage of research
How do Librarians and data services get involved throughout the data lifecycle?
We’re working of a triage model
Liaison librarians are taking great initiative and undergoing training to handle day to day data needs such as DMP consultations, lab protocols, finding disciplinary repositories, and teaching data best practices.
More challenging or novel data issues will trickle up to our data team in the library including a metadata and repository specialist
Truly novel problems that break new ground or requite expertise will be handled by our research data specialists as well as myself and any other future postdoctoral fellows
We’re really working to employ a carrot only approach
We want to enhance research not add any burdens to faculty and students so we’re trying to solve existing pain points in the ressearch workflow
And as part of that happy outreach we’ve decided to use new grant awards as our introduction – so we’ve worked with the office of research to be notified of new grant awards, right now focusing on NIH and NSF, so we can contact the faculty member, congratulate them, and offer our services right at the start (or near the start) of a project.
Like many of you this is still a work in progress – we have a lot more that we want to be able to do, but right now we want to get the word out that we are here, we care about helping campus with data and we’re listening for what researchers want
IMPORTANTLY – we want to be realistic about building our services into the existing research culture
Thanks so much
I’m Stephanie Simms, as a research data specialist at the California Digital Library I think about how to support RDM activities across the 10 UC campuses; my primary responsibility is support for the DMPTool which expands my RDM world to the 173 partner institutions that currently use the tool
I’ve learned that there is no one-size-fits-all solution to providing RDM programs and services
One thread that does tie things together across our panel and other institutions with RDM initiatives is the concept of the User Experience (UX), which is what I want to talk about today
ISO definition of UX: recognizing that people and their experiences matter
This typical Boston traffic sign illustrates the importance of UX: anyone who’s driven there knows that it is an extremely frustrating, unsatisfying experience
I didn’t know that UX was a thing until I began my CLIR Postdoc at UCLA last year and started thinking more critically about the major roadblocks I experienced managing my own research data: top two were education about RDM and easy-to-use tools
Now I feel privileged to work with a professional UX team at the California Digital Library who enhance my appreciation of user-centered design on a daily basis
It may seem obvious, but once you shift your focus and think of things from the user’s perspective the value of building websites, services, and training materials that are intuitive and easy to use becomes clear
In order to be effective, both “training up” and “reaching out” efforts for RDM need to consider UX, for all stakeholders but especially for library staff offering RDM services and for researchers who are ultimately the ones responsible for managing their research data
First impressions are everything – as we learned in the keynote yesterday, most people rely on internet searches as the primary means of accessing information
Research done by the Harvard UX library reveals that LibGuides are not the best way to discover info and make that first impression
These excellent resources are not easy to find (buried in mass of library web content), if you manage to get to this specific page it’s not easy to find the information you’re looking for
Contrast the LibGuide swirl of text and links with this website, which is beautifully designed by Brianna Marshall for the Univ of Wisconsin Research Data Services
Easy to find what you’re looking for, inspires confidence, it makes it easy for researchers there to “get in touch” thanks to this intuitive red button to get more info about RDM
I also want to point out that nowhere on this website will you find the overwhelming, and ever-clockwise, research data lifecycle diagram – these have their place, but it’s probably not the homepage of an RDM website
Now I want to mention a couple of simple UX methods that are useful not just for designing websites but for designing other training and outreach materials, and when considering communication strategies about RDM across campus
Personas: represent the goals and behavior of a hypothesized group of users. Captured in 1-2 page descriptions.
DataONE (NSF-funded initiative for managing environmental science data) recently released a batch of personas, here are just 2 that consider role-based needs for RDM
By developing personas you can tailor info and services for diverse audiences, with the goal to not overload people with extraneous info,
also makes RDM relevant to their work –
Another obvious idea: that in order for a thing to be effective, you have to design it for a specific person
Usability testing provides information about whether a person can effectively and efficiently do something
Building some form of assessment into services (whether formal or informal) allows libraries to take an evolutionary approach and continuously adapt to constantly shifting RDM landscape and as they learn about needs across campus
you don’t need large sample sizes to begin identifying stumbling blocks
An informal usability test of this door with just 1 or 2 people should illuminate some usability problems
Usability is main driver for a year-old service from the California Digital Library called Dash. Dash is an online portal designed to make campus data sharing easy. While Dash gives its users the impression of being a full-fledged repository, in actually it is a lightweight overlay layer that sits on top of an existing repository available to the UCs and hides other processes from view (assigning a DOI, CC-BY license, minimal metadata reqs).
Files can be deposited by drag-and-drop or browsing an attached file system. We’re halfway through a grant-funded project to refactor the UI to make it even easier to use.
Many libraries are venturing into new RDM territory and need to secure collective buy-in from diverse stakeholders across campus to be successful
To avoid cognitive dissonance, it’s important to think about easy to use and rewarding approaches
That way we can avoid frustration, improve RDM experience for everyone