Data Management for Research

Report
Data Management
for Research
Aaron Collie, MSU Libraries
Lisa Schmidt, University Archives
Introductions
 Please tell us your name and
department
 A brief description of your
primary research area
 What do you consider to be your
research data?
 Optional:
 Experience managing research
data?
 Experience writing a data
management plan?
cc http://www.flickr.com/photos/quinnanya/
Agenda
• Introductions
• Background
• Definitions
• Upfront Decisions
• Data Sharing Impacts
• Fundamentals Practices
• File Organization
• Data Documentation
• Reliable Backup
• Data Lifecycle Strategy
Why are we here?
But why are we really here?
 An Impetus: NSF recently released a mandate that all grant
applications submitted after January 18th, 2011 must include a
supplemental “Data Management Plan”
 An Effect: This mandate from NSF has had a domino effect,
and many funders that now require or state guidelines for
data management of grant funded research
 A Challenge: Data management (and oftentimes research
methods in general) is an area that has not traditionally
received a full treatment in most graduate and doctoral
curricula
What is meant by “data management”?
Fundamental Practices
 File Organization
 Data Documentation
 Reliable Backups
Data lifecycle




Digital Sustainability
Scholarly
Communication
Data Publishing
Research Impact




Effective January 18, 2011
NSF will not evaluate any proposal missing a DMP
May be up to two pages long
PI may state that project will not generate data or
samples
 DMP is reviewed as part of intellectual merit or
broader impacts of application, or both
 Costs to implement DMP may be included in
proposal’s budget
NSF’s Data Management Guidelines
 Policies for re-use, re-distribution, and creation of
derivatives
 Plans for archiving data, samples, and other research
outcomes, maintaining access
 Types of data, samples, physical collections, software
generated
 Standards for data and metadata format and content
 Access and sharing policies, with stipulations for
privacy, confidentiality, security, intellectual property,
or other rights or requirements
Other Federal Policies
“expects the timely release and sharing of final research
data"
“…should describe how the project team will manage
and disseminate data generated by the project”
“requires that data…be submitted to and archived by
designated national data centers.”
NASA “promotes the full and open sharing of all data”
"IMLS encourages sharing of research data."
Upfront Decisions for Researchers
 What is the expected lifespan of the data?
 Besides the researcher(s) on the project, who else
should be given access to the data?
 Does the dataset include any sensitive information?
 Who owns or controls the research data?
 Should any restrictions be placed on the dataset?
 How are the data stored and preserved?
Upfront Decisions for Researchers
 How might the data be used, reused, and
repurposed?
 How is the data described and organized?
 Who are the expected and potential audiences for
the datasets?
 What publications or discoveries have resulted from
the datasets?
 How should the data be made accessible?
Data Sharing Impacts
 Reinforces open scientific
inquiry
 Encourages diversity of
analysis and opinion
 Promotes new research,
testing of new or
alternative hypotheses
and methods of analysis
 Supports studies on data
collection methods and
measurement
Cc http://www.flickr.com/photos/pinchof_10/
Data Sharing Impacts (cont.)
 Facilitates education of
new researchers
 Enables exploration of
topics not envisioned
by initial investigators
 Permits creation of
new datasets by
combining data from
multiple sources
Agenda
• Introductions
• Background
• Definitions
• Upfront Decisions
• Data Sharing Impacts
• Fundamentals Practices
• File Organization
• Data Documentation
• Reliable Backup
• Data Lifecycle Strategy
File Organization Practices: Overview
1. Create a file plan for your
research project
2. Design a file naming
convention that works for
your project
3. Agree on a version control
method to assist with file
synchronization
4. Carefully choose file
formats to maximize
usefulness
“When I was a
freshmen I named
my assignments
Paper Paperr
Paperrr Paperrrr”
-Undergrad
1. Create a file plan for your research project
 File plan as a classification system
 Indexed – makes it easier to locate folders/files
 Primary subjects – main functions of research project
 Secondary subjects – more specific activities of project,
including research data
• Tertiary subjects – limit by date or equivalent
– File Name (naming conventions)
1. Create a file plan for your research project
(cont.)
Example documentation of Directory Hierarchy:
 /[Project]/[Grant Number]/[Event]/[Date]
Example documentation of File Naming Convention:
 [investigator]_[method]_[descriptor]_[YYYYMMDD]_[version].[ext]
2. Design a file naming convention that works
for your project
 Why file naming conventions?
 Enable better access/retrieval of files
 Create logical sequences for file sorting
 More easily identify what you’re searching for
2. Design a file naming convention that works
for your project (cont.)
 Meaningful but short (255 character limit)
 Descriptive while still making sense
 Capital letters or underscores differentiate
between words
 Surname first followed by initials of first name
 More on handout
2. Design a file naming convention that works
for your project (cont.)
This
Not This
sharpeW_krillMicrograph_backscatter3_20110117.tif
KrillData2011.tif
This
Not This
borgesJ_collocation_20080414.xml
Borges_Textbase.xml
3. Agree on a version control method to assist
with file synchronization
 Version number of record indicated file name
with “v” followed by version number
 Letter “d” indicates draft
Examples of simple version control:
waltM_lakeLansing_fieldNotes_20091012_v002.doc
petersK_OrgChart2009_d001.svg
4. Carefully choose file formats to maximize
usefulness
•
•
•
•
•
•
Non-proprietary
Open, documented standard
Common usage by research community
Standard representation (ASCII, Unicode)
Unencrypted
Uncompressed
Documentation Practices: Overview
1. At minimum create a
README file that you can
use to document your
project
2. Utilize standards for
describing data including
Metadata Standards
3. If applicable, use in-line
code commentary to
explain code
(cc) Will Scullin
1. At minimum create a README file that you
can use to document your project
 At minimum, store documentation in readme.txt file or
equivalent, with data
 Resource: http://libraries.mit.edu/guides/subjects/datamanagement/metadata.html
2. Utilize standards for describing data including
Metadata Standards
 “Data about data”
 Standardized way of describing data
 Explains who, what, where, when of data
creation and methods of use
 Provides the essential tools for discovery, such as
a bibliographic citation
2. Utilize standards for describing data including
Metadata Standards
Basic project metadata:
• Title
• Language
• File Formats
• Creator
• Dates
• File Structure
• Identifier
• Location
• Variable List
• Subject
• Methodology
• Code Lists
• Funders
• Data Processing
• Versions
• Rights
• Sources
• Checksums
• Access
Information
• List of File Names
Documentation Practices: Example Metadata Standards
 Dublin Core
Easy-to-create-and-maintain descriptive format to
facilitate cross-domain resource discovery on the Web
 Darwin Core
Facilitates reference and sharing of biological diversity
datasets
 Data Documentation Initiative (DDI)
Methodology for content, presentation, transport, and
preservation of metadata about datasets in the social
and behavioral sciences
Documentation Practices: Example Metadata Standards
 Directory Interchange Format
Descriptive format for exchanging information about
earth science data
 ISO 19115:2003
Describes geographic data such as maps and charts
 PBCore
Supports description and exchange of media assets,
including both individual clips and full, edited, aired
productions
Documentation Practices: Example Metadata Standards
 Science Data Literacy Project
Metadata for astronomy, biology, ecology and
oceanography
 VRACore
Data standard for description of works of visual culture
as well as images that document them
3. If applicable, use in-line code commentary to
explain code
Example of R code commentary
# Cumulative normal density
pnorm(c(-1.96,0,1.96))
Backup Practices: Overview
1.
2.
3.
4.
Avoid single points of failure
Understand the different types of storage
Ensure data redundancy
Aim for geographic distribution of data
1. Avoid single points of failure
A single point of failure occurs when it would only take one
event to destroy all data on a device (e.g. dropped hard drive)
Good practices for avoiding single points of error:
 Use managed networked storage whenever possible
 Move data off of portable media
 Never rely on one copy of data
 Do not rely on CD or DVD copies to be readable
 Be wary of software lifespans (e.g. Angel)
2. Understand the different types of storage
•
•
•
•
•
•
Flash Drives
Internal Hard Drives
External Hard Drives
Server and Web Storage
Managed Networked Storage
Cloud Storage
3. Ensure data redundancy
Backup Do’s:
 Make 3 copies
 E.g. original + external/local + external/remote
 E.g. original + 2 formats on 2 drives in 2 locations
 Geographically distribute and secure
 Local vs. remote, depending on needed recovery time
 Personal computer, external hard drives,
departmental, or university servers may be used
3. Ensure data redundancy (cont.)
Backup Don’ts:
 Do not rely on one copy
 Do not use CDs and DVDs
 Do not rely on ANGEL
(cc) George Ornbo
3. Ensure data redundancy (cont.)
Backup Maybe:
 Cloud storage





Note that many
enterprise cloud
storage services
include a charge for
in/out of data
transfers
Amazon s3
Google
MS Azure
DuraCloud
Rackspace
$$$
Agenda
• Introductions
• Background
• Definitions
• Upfront Decisions
• Data Sharing Impacts
• Fundamentals Practices
• File Organization
• Data Documentation
• Reliable Backup
• Data Lifecycle Strategy
Research is…
?
The scientific method “is
often misrepresented as a
fixed sequence of steps,”
rather than being seen for
what it truly is, “a highly
variable and creative
process” (AAAS 2000:18).
Gauch, Hugh G. Scientific Method in Practice. New York: Cambridge University Press, 2010. Print. (Emphasis added)
Scientific Method
More Generic
The Research Depth Chart
Research Method
Research Tasks
More Specific
Research Design
The Data Management Depth Chart
Research Data Lifecycle Model
Source: DDI Structural Reform Group. “Overview of the DDI Version 3.0 Conceptual Model.“ DDI Alliance. 2004.
http://opendatafoundation.org/ddi/srg/Papers/DDIModel_v_4.pdf
The Data Management Depth Chart
Research Data Lifecycle Model
???
???
Research Data Management Tasks
The Data Management Depth Chart
Research Data Lifecycle Model
Data Management Plan
???
Research Data Management Tasks
 http://www.lib.msu.edu/about/diginfo/ldmp.jsp
Data are brainstormed
Study
Concept
Data are brainstormed
DMP
• Data type, purpose & value
MSU
• University Research Council guidelines
• Research Facilitation and
Dissemination
• Lifecycle Data Management Planning
• Research Data Management Guidance
YOU
• Start your Data Management Plan!
Data are collected and secured
Study
Concept
Data
Collection
Data are collected
DMP
MSU
YOU
• Data format, size & short term storage
• ATS Andrew File System (AFS)
• Institute for Cyber Enabled Research
• MSU Libraries Data Services
• MSU Libraries Campus Data Resources
• File Plan, File Naming, Backup Plan
Data are normalized and processed
Study
Concept
Data
Collection
Data
Processing
Data are processed
DMP
MSU
YOU
• Data transformations & structures
• LCT Computing Courses
• High Performance Computing Center
• Consortium of Research Consulting
Services
• Documentation, Methodology
Data are distributed
Study
Concept
Data
Collection
Data
Processing
Data
Distribution
Data are distributed
DMP
• Data sharing, security & rights
MSU
• Human Research Protection Program
• University Research Council guidelines
• MSU Libraries Copyright Permissions
Center
• MSU Google Apps
YOU
• Roles, Responsibilities, Resources
Data are discoverable
Study
Concept
Data
Collection
Data
Processing
Data
Distribution
Data
Discovery
Data are discoverable
DMP
• Data publishing & metadata
MSU
• Development of Copyrighted Materials
• MSU Libraries Data Citation Guide
YOU
• README, Metadata Standard
Data are analyzed
Study
Concept
Data
Collection
Data
Processing
Data
Distributio
n
Data
Discovery
Data
Analysis
Data are analyzed
DMP
• Standards & workflow documentation
MSU
• Center for Statistical Training and
Consulting
• Statistical Consulting Services
YOU
• Code Commentary, Documentation
Data are stored and preserved
Data
Archiving
Study
Concept
Data
Collection
Data
Processing
Data
Distribution
Data
Discovery
Data
Analysis
Data are preserved
DMP
• Long term storage & management
MSU
• VPRGS Repositories and Archives
• Lifecycle Data Management Planning
• Databib.org!
YOU
• Embrace stewardship
Data can be used and reused
Data
Archiving
Study
Concept
Data
Collection
Data
Processing
Data
Distribution
Data
Discovery
Repurposing
Data
Analysis
Data can be used and reused
DMP
• Broader impact
MSU
• Research Data Management CAFE
• MSU Research Centers and Institutes
• MSU Libraries Data Citation Guide
YOU
• Publish your data!
Research Data Management Guidance
 Face-to-face Advising
 Writing Data Management Plans
 Planning for Digital Projects
 Managing Digital Information
 Group Training
 New Faculty Orientation
 Faculty Seminars
 Classroom Instruction
lib.msu.edu/about/rdmg
In Conclusion…




Upfront Decisions Researchers Need to Make
General Good Practices for Managing Research Data
NSF, NIH, IMLS and Other Funders’ Requirements
Lifecycle of Research Data
Contact
Lisa M. Schmidt
Electronic Records Archivist
University Archives & Historical Collections
[email protected]
Aaron Collie
Digital Curation Librarian
MSU Libraries
[email protected]

similar documents