Introduction to Research Data Management for postgraduate students

Introduction to Research Data Management
For postgraduate Students

University of Northampton, 20th February 2013

Marieke Guy
DCC, University of Bath
m.guy@ukoln.ac.uk
Funded by:
This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 2.5 UK: Scotland
License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/2.5/scotland/ ; or,
(b) send a letter to Creative Commons, 543 Howard Street, 5th Floor, San Francisco, California, 94105, USA.

Today’s Talk…

Will consider:
• What are data
• What is the research process
• What is research data management and why does it matter
• How you can start thinking about good research data
management
We hope you leave able to explain why research data
management is important and what roles postgraduate
students / researchers play.

What are data?
• The lowest level of abstraction from which information and
knowledge are derived
• Research data are collected, observed or created, for the
purposes of analysis to produce and validate original
research results
• Both analogue and digital materials are 'data'

• Digital data can be:
• created in a digital form ("born digital")
• converted to a digital form (digitised)

What is the research process?

Research
Process

•Research360

What is research data management?
• Data curation is “the active
management and appraisal
of data
• Over the lifecycle of
scholarly and scientific
interest”
• Data have importance as
the evidential base
• of scholarly conclusions
• Curation is part of good
research practice

Why carry out RDM?
Managing your data properly:
Sharing data can
• Can prevent data loss
increase citation
• Makes researching easier
• Helps with validation of results and
research integrity
• Offers new research opportunities and
collaborations
• Is inline with the UON policy released in
June 2011 More citations: 69% ↑

• Is important for UON’s reputation (Piwowar, 2007 in PLoS)

• Is part of good research practice

Data lifecycle
1. What data will you
produce?
5.
Preservation
1. 2. How will you organise the
Create
& Re-Use
data?
3. Can you/others understand
the data
4.
2.
Publication
& Deposit
Active Use
4. What data will be deposited
and where?
3.
Documentation 5. Who will be interested in
re-using the data?

Conceptualise and plan

Activities
• define a research question and design your methodology
• bid for funding (incl. data management and sharing plans)
• plan data creation (capture methods, standards, formats)
• data management plans!

Roles
PGR student, supervisory team, sponsors / funding bodies, IT,
research governance, ethics panel

Decisions made now have an impact on every other stage of the
lifecycle, so it is worth getting things right from the start!

1. What data will you produce?

• What type of data will
5.
you produce?
1.
Preservation
Create
& Re-Use
• What types of file
format?

4.
• How easy is it to create or
2.
Publication
& Deposit
Active Use reproduce?

• Who owns it and is
3.
Documentation responsible for it?

Data can take many forms
• Notebooks & lab books
• Instrument measurements
• Experimental observations
•http://www.flickr.com/photos/charleswelch/3
597432481//
• Still images, video & audio
• Consent forms •http://www.google.co.uk/imgres?
q=illumina+bgi&hl=en&client=firefox-
a&hs=Jl2&rls=org.mozilla:en-

• Text corpuses GB:official&biw=1366&bih

• Models & software
• Survey results & interview transcripts

Data types
Data Type Value Example
Observational data Usually irreplaceable Sensor readings,
captured around the time telemetry, survey results,
of the event neuro-images
Experimental data from Often reproducible but can Gene sequence,
lab equipment be expensive chromatograms, toroid
magnetic field readings
Simulation data generated Model and metadata Climate models, economic
from test models (inputs) more important models
than output data.

Large modules can take a
lot of computer time to
reproduce
Derived or compiled data Reproducible Text and data mining,
(but very expensive) compiled databases, 3D
models

Who owns or is responsible for
your data?
Ownership
• Data ownership is complex, often defined on a case-by-case
basis
• May be dependent on individual contractual agreements
• Contracts define needs of the University, staff, students,
funders, collaborators

Management
• Contact the University of Northampton ethics department

Responsibilities for data
In practice - Everyone plays their part

If you’re generating and using data, you should:
• Comply with guidelines from your group, department, faculty, collaborators
• Make sure your data is securely stored and backed up
• Describe your data so that you/others can understand it in future

If you’re managing a project, you should:
• Be fully aware of funder, collaborator and publisher requirements
• Ensure you have access to group data
• Assess what should be published and/or archived

• More info: http://www.data-archive.ac.uk/create-manage

2. How will you look after your data?

5.
1.
Preservation
& Re-Use
Create
• Is your data safe?

• Is your data organised?
4.
Publication
2.
Active Use • Can you find your data?
& Deposit

3.
Documentation

Storage and Security
3… 2… 1… Backup!
at least 3 copies of a file
on at least 2 different media
with at least 1 offsite

Test file recovery
At set up time and on a regular basis

Access
Protect your hardware
If sensitive use file encryption
Keep passwords safe (e.g. Keypass)
At least 2 people should have access to your data
More info: http://www.data-archive.ac.uk/create-manage/storage

Storage & Security – Back up options
Media Advantages Disadvantages
CDs or DVDs • Useful for quick restore in • Static capture of data
the event of minor disaster • Not built to last
• Vulnerable to theft
• Physical loss of media
External hard • Dynamic capture of data • Must store securely and
drives, including • Useful for quick restore in remotely to original copy
dropbox the event of minor disaster • Vulnerable to theft
• Must use file encryption if
sensitive
Northampton • Resilient backup • Lack of offline access
server • Must have a Northampton
• TUNDRA2 will replace account
Digital scans of • Can carry out using campus • Manipulation of page
lab books printer content difficult

Can you find your data?
A Clear Directory Structure
• Top level folder and substructure
File Version Control
• Discard obsolete versions if no longer needed after making
backups
• Manage using: File naming (see below), version control software
(e.g. Git, Mercurial, SVN)
File Naming Conventions
• Record any naming conventions or abbreviations used e.g.
[Experiment]_[Reagent]_[Instrument]_[YYYYMMDD].dat
• Date/time stamp or use a separate ID (e.g. v1) for each version
More info: http://www.jiscdigitalmedia.ac.uk/crossmedia/advice/
choosing-a-file-name/

3. Documenting data

• Do you still understand
your older work?
5.
1.
Preservation
Create
& Re-Use
• Is the file structure /
naming understandable
to others?
4.
2.
Publication
& Deposit
Active Use
• Which data will be kept?

3. • Which data can be
Documentation
discarded?

Understanding your data
• Students:
• Will you be able to write up your methods at the end of
your studies?
• Project leads:
• Will you be able to respond to reviewers comments?
• Will you be able to find the information you need for final
project reports?
• Can you reproduce your work if you need to?
• What information would someone else need to replicate
your work?

Understanding your data
Do you know how you generated your data?
• Equipment or software used
• Experimental protocol
• Other things included in (e.g.) a lab notebook
• Can reference a published article, if it covers everything
Are you able to give credit to external sources of data?
• Include details of where the data are held, identified & accessed
• Cite a publication describing the data
• Cite the data itself e.g.

Metadata
• Contextual information for data is called metadata — literally
data about data
• Data repositories & archives require some generic metadata,
e.g. author, title, publication date
• For data to be useful, it will also need subject-specific
metadata e.g. reagent names, experimental conditions,
population demographic
• Record contextual information in a text file (such as a ‘read
me’ file) in the same directory as the data e.g.
• codes for categorical survey responses
• ‘999 indicates a dummy value in the data’
More info: http://www.data-archive.ac.uk/create-manage/document

3.What data will be deposited
& where?
• Are you expected to
share your data?
5.
1.
Preservation
& Re-Use
Create
• Are you allowed to
share your data?

• Define the core data set
4.
Publication
2.
Active Use
of the project
& Deposit

• Which data will be
3.
Documentation included in your
publication / thesis?

Data Sharing – Why share your
data?
• Share with your future self – avoid repeating research!

• Promote your research – get cited!

• Enable new discoveries

• Replication

• Store your data in a reliable archive

• Comply with funding requirements

Requirements to share your data
• Some journal publishers have a policy on data availability.
• Most UK funders now expect research data to be made
publically available – UON policy
• Are you making any of your data available as supplementary
information?
• Is there sufficient information with the data so that it can be
understood and reused?

declaration
data are a public good and Code of good research conduct
should be openly available data should be preserved and accessible
for 10 years +
Funders’ data policies Common principles on data policy
www.dcc.ac.uk/resources/policy-and- www.rcuk.ac.uk/research/Pages/
legal/funders-data-policies DataPolicy.aspx

Restrictions on sharing your data

Are there privacy requirements from the funders or commercial
partners? e.g. personal data, high security data
• You might not have the right to share data collected from
other sources
• It depends upon whether those data were licensed and
have terms of use
• Most databases are licensed and prohibit redistribution of
data without permission
• If you are uncertain as to your rights to disseminate data,
check with the ethics department

How to share your data
• Deposit in a data repository eg. GenBank, UKDA
• Data can be licensed
• Culture of data sharing: can make available your data under a CC-
BY or CC0 declaration to make this explicit
• CC-BY license permits reuse but requires attribution
• CC0 declaration is a waiver of copyright.
• Note that laws about data vary in different countries.

• You may have rights to first use or to commercial exploit data
How to license research data:
http://www.dcc.ac.uk/resources/how-guides/license-research-

5. Preservation and reuse

• How long will your data
5.
be reusable for?
1.
Preservation
Create
& Re-Use
• Do you need to prepare
your data for long term
archive?
4.
2.
Publication
Active Use
& Deposit
• Which data do you need
to keep?
3.
Documentation

Data retention and archiving
How permanent are the data?
• Short term (e.g. 3-5 years)
• Long term (e.g. 10 years)
• Indefinite

Should discarded data be destroyed?
• Keep all versions? Just final version? First and last?

What are the re-processing costs?
• Keep only software and protocol/methodology information

Are there tools/software needed to create, process or visualise
the data? Archive these with your data

Selection and appraisal
 Make a start on selection and appraisal from as early a point as possible.

 Plan for what you think you’ll need to keep to support your research findings. What is
the minimum you’ll need to support your findings over time?
 Know who you are keeping it the data for and what you want them to be able do with
it (is for yourself only, or for other people too?). This may affect the way you keep it
and what you keep.
 Conversely, know what you need to dispose of. Destruction is often vital to ensure
compliance with legal requirements.
 Appraise for the here and now but with an eye to the future.

 Think about resources required / available. These will affect you selection and
appraisal decisions.
 Look for relationships with other data sets in your archive/repository as part of the
appraisal process (i.e. does the dataset augment another collection significantly?).
 Some funders stipulate that you must identify whether the data exists already. This
process might highlight additional datasets that your new research might augment
significantly.

File formats for long term access
• Unencrypted
• Uncompressed
• Non-proprietary/patent-encumbered
• Open, documented standard
• Standard representation (ASCII, Unicode)
Type Recommended Avoid for data sharing
Tabular data CSV, TSV, SPSS portable Excel
Text Plain text, HTML, RTF Word
PDF/A only if layout matters
Media Container: MP4, Ogg Quicktime
Codec: Theora, Dirac, FLAC H264
Images TIFF, JPEG2000, PNG GIF, JPG
Structured data XML, RDF RDBMS

Further examples: http://www.data-archive.ac.uk/create-manage/format/formats-table

Summary
• Data management is important at all stages of a project
• There are tools available to help you
• Keep your data safe: Back up your data and test your back-
ups
• Keep your data organised
• Find it – good formats and file names
• Understand it - check documentation and metadata
• Consider publishing your data so that you can get recognition
for your work
• Help is available at:
http://researchsupporthub.northampton.ac.uk/contact-us/

Thanks - any questions?

Acknowledgements:
Thanks to Research360, DCC staff, UK Data Archive, Mantra
(University of Edinburgh) for slides

Introduction to Research Data Management for postgraduate students

More Related Content

What's hot (20)

Viewers also liked (6)

Similar to Introduction to Research Data Management for postgraduate students (20)

More from Marieke Guy (20)

Recently uploaded (20)

Introduction to Research Data Management for postgraduate students