DataONE All Hands Meeting 10/22-24/2013 - PPSR working group (Todd Suomela, Chris Eaker and Laura Moyers representing UT present)
12:30-3:00
Jennifer Shirk (CLO), Andrea Wiggins (CLO), Anne Bowser (CLO-summer intern), Robert Stevenson (UMass), Greg Newman (co-chair, CSU-citsci.org), Megan Hines (USGS), Julian Turner (CSU-CoCoRAHS), Todd Suomela (UTK), Laura Moyers (UTK), Rick Bonney (co-chair, ), Chris Eaker (UTK-Hodges Library, Data Curation) briefly
Primary topic of conversation: Redesign of citsci.org, specifically the PPSR-CORE metadata. This is project-level metadata; project will have a GUID. The project's datasets' metadata will point to this project-level metadata (i.e. include the project GUID in each dataset's metadata)
Plan to devote most of Thurs/Fri to writing
Today - Greg and Jennifer's show - citsci.org and how it relates to citizen science in general, Ideum guys (Ben and Miles - building next citsci)
metadata and data sharing issues to resolve, identify who's doing what where
Jennifer: citizen science exploding, 1000s of projects; need to understand the field, networking, in-depth research either about the field or about a given project in context of the field; USGS (and other federal entities) has list of PPSR projects;
project-level metadata - project name, desc, URL, # participants, etc. -- want simple to start with
scistarter.org - http://scistarter.com/index.html
citsci.org - http://citsci.org/cwis438/websites/citsci/home.php?WebSiteID=7 (Greg Newman)
informalscience.org - http://informalscience.org/
cocorahs.org - http://cocorahs.org/ Community Collaborative Rain, Hail & Snow Network, Colorado State (Julian Turner)
citizensciencecentral.org - at Cornell - http://www.birds.cornell.edu/citscitoolkit
need a metadata architecture (user data about citizen science projects) that crosses all PPSR entities ---- PPSR_CORE data standard
also, need a dashboard - how many PPSR projects are out there, what kinds of data to they hold, what is the impact of citsci, networking,
spatial/temporal data highly desired but problematic, number of papers, citations, etc.
Miles: is it a good idea to have one metadata structure that all PPSRs adhere to
informalscience.org is like a dataone for PPSR
goal for today: define fields (does this mean common metadata standard)
want "most recent data" capability
PPSR has living data sets
Discussion of proposed REQUIRED FIELDS:
Miles showed a spreadsheet PPSR_CORE_v2
- ProjectGUID - GUID for each PPSR project
- owner of project needs to generate a GUID or ask someone like citsci.org to do it for them
- ProjectName - name of project
- ProjectDescription - goals, objectives, purpose, vision, etc. (there are different types of project descriptions, i.e. public-facing, for participants, for funders, for data users, etc.) a description that is universal; no minimum/maximum
- can add additional description fields for entities which require different/additional information (not required)
- ProjectDateLastUpdated - date the project information (this metadata) was last updated
- ProjectCoordinatorName / ProjectContact / ProjectContactEmail - discussion about making contacts separate (i.e. not in-line in this table), what if you have one project in citsci, then later you have a diff project in citizenscience, how will it know if you are the same person; Coordinator and Contact are often two different people - DECISION: keep ProjectContact which links to a separate table with name, email, phone, etc. Keep ProjectContactEmail as non-required - if it is omitted, pull email from ProjectContact record in separate table
- ProjectStatus - "Starting" or "Active" or "Complete" or "Hiatus" or "Pending" <-- controlled vocab for these statuses is desirable. Important so that we don't contact people whose project is closed.
- ProjectStartDate / ProjectEnddate are optional fields
- OPTIONAL for further discussion: ProjectNumParticipants, ProjectNumActiveParticipants (lots of confusion from Andrea's survey, what does "registered" mean, or contributor, or active, etc.) Important for use case studies, but very difficult to define. registered vs active, all-time vs current
maybe citsci.org could be like LTER - a network of networks, exposing metadata for all these other PPSR entities.
3 models of cit sci project owners
no website - need someone else to generate GUID and/or store data
CoCoRAHS - capable of generating GUID
citsci.org and others - serve as GUID providers / data storers
Laura needs to attend later meetings on this topic to follow up.
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
10/22/13 3:30-5:00
Ideum plans to draft up XML and mock-up of webform to enter data and get back with PPSR
Scheduling of efforts - keep things moving along, but need community feedback
Dashboard - what kind of metrics are desired - this can drive the kinds of metadata you want to require "Begin with the end in mind"
Further discussion on desired metadata fields, especially regarding later reporting needs
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
10/23/13 8:30-10:00
Needs if PPSR entities want to be MN or somehow contribute to an existing MN
- Decision on how often to take a snapshot of data
- yearly, quarterly, monthly
- How to handle corrections (QC)
- How to handle updates of data
- append "current year" file, separate file "date range", etc.
- What is citsci.org thinking about doing? Are they a "network of networks" like KNB and LTER? Greg says they host data for PPSR projects as well as metadata.
- CoCoRAHS (Julian) is very interested in slender node. <--- get with Dave re: slender node and streaming/live data
- See the PPSR MN persona - eBird is different from Zooniverse in terms of data types, data collection, etc.
CAISE (Center for Advancement of Informal Science Education) c.f. informalscience.org, see 2009 paper "Public Participation in Scientific Research: Defining the Field and Assessing Its Potential for Informatl Science Education" http://eric.ed.gov/?id=ED519688 (focus on educational outcomes)
See PPSR WG charter
See the Tool Kit / steps on http://www.birds.cornell.edu/citscitoolkit/toolkit
Need an ongoing PPSR group, beyond DataONE, see NSF and others' emphasis on citizen science (LM)
3 C's model (co-created, contributory,
rubric for scientific outcomes
scientists very concerned with data quality
Rick proposed a white paper focused on scientific outcomes of citizen science, case studies
Rob suggests a (separate) paper with emphasis on data quality
See How Science Works: http://undsci.berkeley.edu/article/howscienceworks_01 <--- share this with Bruce et. al.
Investigate PPSR WG's reputation within DataONE and in the broader PPSR community (LM)
Using humans as sensors is crowdsourcing, faux-data collection as an educational tool isn't "real" either. Need both good scientific outcomes and citizen participation for it to be real citizen science.
Julian is really into metadata (value of good metadata in data discovery, etc.)
Requests for participation from data users is much bettter received than requests from repository (my time's not wasted, etc.) Julian's "data story"
Andrea - can get funding for education rather than science, even if science is the ultimate goal.
see "biscuit paper" as a stepping stone to a PPSR directorate in NSF (see line 88 above) woo hoo!!! (Bill said when finished he and Rick could present to NSF)
Zooniverse (massive data collection, papers), case study for "biscuit paper"
If we want the steps to be useful to the community, we need to demonstrate their utility. Factors, rather than linear steps. A set of (not exclusive) factors.
See foldit project
What is 4th paradigm (data-driven research rather than working from a hypothesis)
2nd morning session: plan to work as group on the proposed "biscuit" paper, Megan to take notes.
=============================================================================================================
10/24/13, 10:30-noon breakout for data characteristics that drive good science outcomes
What are the things you can do to elevate trust/usefulness in/of data?
Julian, Rob, Megan
Looking at it from the user/data searcher's perspective
Sometimes difficult to find ALL the data for a PPSR project. Sometimes a PPSR project will only make available a portion of their data because ALL the data is overwhelming or not useful to a majority of users or
discoverability - ensuring data is visible to users, can they be discovered via an internet search engine, is the dataset registered in a repository (specific to scientific domain), are common terms used to describe the data (metadata standards?),
accessibility/usability - "the extent to which information is available, or easily or quickly accessible" - see Hunter et al 2012 paper (Megan has) can you get the data in a raw format? do you have an API? is it publicly available?
trust - is project/entity trusted/reliable? is data reliable/trusted? consider observational error - how does that impact reliability/believability/trust of data?
=============================================================================================================
10/24/13, 1p-3p, entire group reconvenes to discuss science outcomes paper outline
citsci.org is an infrastructure, not a project (see how funding questions apply)
See Data Exchange Protocol - core fields for project metadata (PPSR_CORE)