Embed Size (px)
Citation preview
Impactful Biomedical Research: Achieving Quality and Transparency
Research Data Management
Laure Perrier
Research Data Management Librarian
June 10, 2016
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Objectives
To describe the current issues facing
researchers with regards to research data
management
To outline strategies for researchers to meet
current challenges
To provide examples of meaningful support
related to research data management
2
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Current Issues
Drivers for research data management
Research funder policies: Encourage or
mandate creating data management plans,
deposit data in repositories
Journals: Require datasets to be published
or made accessible (BMJ, PLOS)
3
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Current Issues
Drivers for research data management
Recent activity,
International Committee of Medical Journal Editors: Sharing Clinical Trial Data (January 2016)
4
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Current Issues
Current Obligations
CIHR
bioinformatics, atomic, and molecular
coordinate data
retain original datasets (all data) for a
minimum of 5 years after the end of the
grant
5
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Current Issues
Future Directions
CIHR
Draft Tri-Agency Statement on Digital Data
Management
– Data management planning
– Preservation, retention, and sharing
6
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Current Issues
7
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Current Issues
Why keep data / make data available?
Find and understand data when needed
Validate results
Ensure research is visible and has impact
Get credit when others cite work
Avoid unnecessary duplication
8
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Current Issues
64 RCTs (oldest at top)
1987 – 2002
Positive findings: aprotininmore effective than comparative treatment
~4,000 participants recruited without need
9
Source: Fergusson D, Glass KC, Hutton B, Shapiro S. Randomized controlled trials of aprotinin in cardiac surgery: could clinical equipoise have stopped the bleeding? Clin Trials. 2005;2(3):218-29
1. Dec-872. Mar-893. Apr-894. Sep-905. Sep-906. Dec-907. Jun-918. Sep-919. Dec-9110. Apr-9211. Jun-9212. Jun-9213. Jun-9214. Nov-9215. Dec-9216. Jan-9317. Jul-9318. Aug-9319. Dec-9320. Jan-9421. Feb-9422. Feb-9423. Feb-9424. Apr-9425. Jul-9426. Aug-9427. Aug-9428. Oct-9429. Oct-9430. Dec-9431. Dec-9432. Feb-9533. Feb-9534. Feb-9535. Apr-9536. Jun-9537. Jun-9538. Sep-9539. Oct-9540. Oct-9541. Oct-9542. May-9643. Jul-9644. Aug-9645. Aug-9646. Oct-9647. Dec-9648. Jan-9749. Jan-9750. Aug-9751. Sep-9752. Dec-9753. Oct-9854. Oct-9855. Nov-9856. Aug-9957. Sep-9958. Mar-0059. Dec-0060. Dec-0061. Jan-0162. Sep-0163. Sep-0164. Jan-02
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies
4 KEY STRATEGIES
10
Make a Plan
Create a System
Secure Your Data
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies
4 KEY STRATEGIES
11
Make a Plan
Create a System
Secure Your Data
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Make A Plan
Data Management Plan
Type of data produced
Documentation (metadata)
Security, storage, management, and back-up of data
Archiving and preservation
Sharing and re-use
12
Make a Plan
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Make A Plan
13
-------------------------------------------------------------------------
Make a Plan
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Make A Plan
14
Make a Plan
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Make A Plan
Tools available for drafting
plans
–DMP Assistant: portagenetwork.ca
–DMP Online: dmponline.dcc.ac.uk
–DMP Tool: dmptool.org
15
Make a Plan
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Make A Plan
Sample Data Management
Plans
–Generic examples (UNC: The
Odum Institute) www.irss.unc.edu/odum/contentSubpage.jsp?nodei
d=570
16
Make a Plan
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies
4 KEY STRATEGIES
17
Make a Plan
Create a System
Secure Your Data
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Strategies: Create a System
18
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Folders: hierarchy
19
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Project Name
Surveys
Instrument 1
Instrument 2
Data
Raw
Processed
Analysis Poster Paper
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Folders: hierarchy
May need to list which files belongin which folders
20
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Folders: piling
21
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Study Poster Paper
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Folders: piling
Less hierarchy = file names needmore detail
22
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Naming Files: Best practices
Avoid special characters (#$%)
Capitals or underscores (FileName.xxx)
Date (ISO): YYYYMMDD
Version information
23
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
24
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Project_YYYYMMDD_ContentDescription_Initials.ext
Project
name
Standardized
date format
Description
of file content
Team
member
identifier
UnderscoreFile
extension
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Naming Files: Use meaningful names
Project/experiment name or acronym
Location/spatial coordinates
Researcher name/initials
Date or date range of experiment
Type of data
25
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Documentation (metadata):
Describes your data set
Data documentation (metadata) helps you
understand data in detail
Helps other researchers find, use, properly
cite your data
26
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Documentation (metadata):
27
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Title
Creator
Dates
Subject
Funders
Rights
Language
Location
Methodology
etc…..
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Create a System
Documentation (metadata):
Many standards for specific research
disciplines
Digital Curation Centre:
www.dcc.ac.uk/resources/metadata-
standards
28
Create a System
Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies
4 KEY STRATEGIES
29
Make a Plan
Create a System
Secure Your Data
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Secure Your Data
30Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Secure Your Data
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Secure Your Data
3-2-1 Rule:
3 copies of your data
2 different locations
More than 1 type of storage hardware
31Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Secure Your Data
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies
4 KEY STRATEGIES
32
Make a Plan
Create a System
Secure Your Data
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data needs to be:
Readable
Accessible
Understandable
33Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data needs to be Readable
Use non-proprietary formats
34Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Yes No
.txt .docx (Word)
.csv .xlsx (Excel)
M4a (MPEG-4) .mov (Quicktime)
.tif .gif or .jpg (images)
XML RDBMS
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data needs to be Accessible
Move data to new media
Average life span ~3-5 years
If no open file format: Preserve software
35Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data needs to be Understandable
Data must include notes
Include details
Others should be able to understand it
36Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data needs to be Understandable
Example: Data Dictionary (quantitative)
37Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Variable Variable Name Variable Type Variable Width Values / Notes
Participant
ID Number
ID Numeric 3 001-900
Date of
Birth
DOB YYYY/MM/DD 1900-2010/1-
12/1-31
Status STAT Numeric 1 1 = alive
2 = deceased
Hemoglobin HB Numeric 2.1 4.0 - 8.0
Urinary
Iodine
UI Numeric 4.1 0.0 – 1000.0
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data needs to be Understandable
Example: Data Dictionary (qualitative)
38Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Code Code Name Explanation
Clarity CLA Coherence of components
Structure STR Arrangement between component parts
Navigation NAV Accurately ascertaining position and
planning for movement through
information
Saliency SAL Quality by which item stands out in
relation to its neighbours
Flow FLO Moving along in a logical, steady manner
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data repositories
Secure, long-term place for
research data
Often can impose appropriate access
restrictions and /or embargoes
39Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data repositories at UToronto
Dataverse
http://dataverse.scholarsportal.info/dvn
TSpace
https://tspace.library.utoronto.ca
Collections UofT (beta)
https://collections.library.utoronto.ca
40Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Strategies: Open File Formats
Data repositories
Subject-specific
–Registry of Research Data Repositories
www.re3data.org (see: Browse)
41Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Open File Formats
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Finally
UToronto
onesearch.library.utoronto.ca/researchdata
42Source: Kristin Briney. Data Management 101 (2015). Retrieved from: http://www.slideshare.net/kbriney/data-management-101-2015
Material in support of a verbal presentation,
not for interpretation as a stand-alone document: May 2016
Questions?
Laure Perrier
43
Impactful Biomedical Research: Achieving Quality and Transparency
Research Data Management
Laure Perrier
Research Data Management Librarian
June 10, 2016
