Research data management during and after your research ; an introduction / Leon Osinski, Maurits van der Graaf

Research data management
during and after your research
an introduction

Workshop PhD students department of
Mechanical Engineering
31-01-2013

Leon Osinski en Maurits van der Graaf
TU/e, IEC / Bibliotheek
Beschikbaar onder een CC-BY licentie

Workshop subject and objectives
Subject
Research data management, or how to deal carefully and honestly with research data
1.
Honestly: research data management in connection with scientific integrity and
questionable practices
2.
Carefully: research data management during your research and after your research

Objectives
1.

2.

Providing insight into the why of research data management during and after your
research;
To gain insight into existing data management practices and needs of PhD students

Workshop program
Program
 13:30-13:35: introduction
 13:35-14:00: presentation “Scientific integrity and research data management during
your research”
 14:00-14:30: discussion
 14:30-14:35: break
 14:35-15:00: presentation “Research data management after your research”
 15:00-15:30: discussion
 15:30-15:40: closure and final discussion topics

Presentation 1
SCIENTIFIC INTEGRITY AND RESEARCH DATA MANAGAMENT
DURING RESEARCH

Scientific integrity and questionable practices
 Deliberately fabricating research data (Schön, Stapel, Fujii)
 Falsificating research data
 Adjusting research data
 Omitting or eliminating unpleasant data, incomplete reporting of results, not
mentioning failed experiments (‘verification bias’), partial reporting of data
(‘cherry picking’)

 Plagiarism: using data from someone else without mentioning the
source

Scientific integrity & research data management
research data management to ensure research integrity and replication

Principle
Presented information is verifiable. Whenever research results are publicized, it is made clear what the data and the
conclusions are based on, where they were derived from and how they can be verified.

Best Practice
• Research must be replicable in order to verify its accuracy. The choice of research question, the research set-up,
the choice of method and the reference to sources studied is accurately documented.
• The quality of data collection, data input, data storage and data processing is guarded closely. All steps taken must
be properly reported and their execution must be properly monitored (lab journals, progress reports,
documentation of arrangements and decisions, etc.).
• Raw research data are stored for at least five years. These data are made available to other scientific practitioners
at request.
• Raw research data are archived in such a way that they can be consulted at a minimum expense of time and
effort.
From: VSNU, The Netherlands code of conduct for scientific practice: principles of good scientific teaching and research (VSNU, 2012), p. 7




>

Research results must be verifiable  research should be replicable  data collection,
storage and processing should be reported and monitored  data management
Untraceable research data: Poldermans denies data fabrication but cannot prove his right
because data are lost

research in progress

Except in view of scientific integrity, during the course of a research
project there will inevitable be a moment when one wishes to store data
safely and accessible for others:






for further use, processing and analysis within one’s own research group
for use in courses
for exchange with researchers elsewhere who collaborate in the project
for assessment by supervisor or tutor
…

research in progress

Why manage your data during your research?





To ensure your research data are accurate, complete, authentic and reliable
To enhance data security and minimise the risk of data loss
To enable others to access your data: sharing data implies data management
To increase your research efficiency: reuse of your data by yourself

Requirements
 Storage (space, equipment)
 Metadata: data documentation in view of transparancy
 Controlled access: who will have access to the data?
[ Dataverse Network: http://thedata.org ; http://www.dataverse.nl/dvn/ ]

Discussion 1
SCIENTIFIC INTEGRITY AND RESEARCH DATA MANAGEMENT
DURING RESEARCH

discussion topics and questions 1

Questions with regard to scientific integrity



What do you do with data sets from failed experiments or wrong measurements? What
criteria are there for establishing failure?
Are all data sets used for publications? What happens to data sets that aren’t used?

Questions with regard to the nature of research data




What is the type of your research data?
• Quantitative measurement data (obtained from lab equipment)
• Observational data (sensor data, neuroimages, survey data)
• Simulation (data derived from test models)
What is your research data format?
• Text: txt, Word, PDF, XML
• Numerical: Excel, SPSS, R
• Multimedia: jpeg, tiff, mpeg, quicktime, vol (3D images)
• Software: Java, C, Matlab, Mathematica
• Discipline specific
• Instrument specific: MRC (cryon-electron micoscropy)


Questions with regard to the research context of your data
 Own, self-generated data or (also) data produced by others?
 Are data produced as part of a collaboration or in the context of an individual
project?

Questions with regard to the management of data
 Estimated volume of data? Sufficient data storage space?
 Who‘s responsible for storage and back up of your data? Yourself, someone
else?
 Who owns or controls your data? Yourself, principal investigator, data
manager? Who decides on access to the data?
 Are any data lost and if so is that a problem?
 Are data selections made? Not everything is to be stored but…?


Questions with regard to accessibility of data during your research
 Are data accessible to others and if yes, to whom and how?
 Are data requested by persons from outside one’s own research group?
 Are there any special privacy or security requirements with regard to your
data?
 Are data documented or described? Are standards used for that?

Presentation 2
RESEARCH DATA MANAGEMENT AFTER YOUR RESEARCH

after your research

Research data management after your research is primarily a matter of
archiving or publishing your data
Why archiving your data after your research?
 Because data providing the evidence for a published paper can be asked for by
others in view of verificating or replicating your results (scientific integrity)
 Because journal, funder or code of conduct demand data to be accessible
 Because data are unique and / or valuable (non-repeatable observations)
 Because data are an asset, worth sharing in order to be reused or built on by
others
UPSIDE: Uniform Principle of Sharing Integral Data and Materials Expeditiously

after your research

Why making your data public (publishing your data)?
 Data are an asset, worth sharing in order to be reused or built on by others
 Data-driven science: building on existing data or combining data to move
science forward or to generate new science (‘Fourth paradigm of science’)

after your research

The best way to share your data is to publish your data with a DOI [digital
object identifier]
Making your data public with a DOI allows you to





claim your data
retrieve your data
cite your data
enhance your publication by integrating data with your paper

after your research

Documentation (metadata)
 If you want to publish your data, mere availability of data has no value. To be
useable, data need to be well-documented
 Data should be made public in an 1accessible, 2intelligible, 3assessable and 4reusable way ‘intelligent openness of data’

after your research

Where to publish your data?
1. Repositories / archives:







3TU.Datacentrum
DANS
PANGEA
Figshare
UCI Machine Learning Repository
…

2. Data journals

Discussion 2
RESEARCH DATA MANAGEMENT AFTER YOUR RESEARCH


Questions
 What happens with your data when your project is finished or when you leave
the university?
 Who owns your data when your project is finished?
 Are data at the end of a project centrally stored?
 How do you ensure your data are easy to understand or analyse?
 Do you use metadata standards to describe your data?


Questions
 Are there any journals within one’s discipline who require data to be made
available?
 Are there research funders within one’s discipline who require data to be made
available?
 Are there examples within one’s discipline of stand-alone data publications?
 Are there reasons not to publish your data, i.e. to make them public?
 Are there privacy, legal or security issues involved with your data?
 Has someone ever asked for your research data?

final discussion topics

 Do you feel a need for more formal data management procedures /
data management plan within your laboratory?
 Would you consider publishing your research data sets?

Research data management during and after your research ; an introduction / Leon Osinski, Maurits van der Graaf

More Related Content

What's hot (20)

Similar to Research data management during and after your research ; an introduction / Leon Osinski, Maurits van der Graaf (20)

More from Leon Osinski (20)

Recently uploaded (20)

Research data management during and after your research ; an introduction / Leon Osinski, Maurits van der Graaf