Archiving and Retention · How to access Research Use Coverage, universe, methodology Time Period: Country: Observation units: Kind of Data: Universe: Data Classification: Time Dimensions:
Post on 08-Oct-2020
2 Views
Preview:
Transcript
Archiving and Retention How can we archive data in Safe Settings?
Yannis KotrotsiosData Access OfficerCancer Research UK
Describing the Problem
ApplicationAccess Data
Personalised or
de-personalised
End of
ContractImpact / Consequences
1. The statistical results cannot be
replicated and independently verified
2. No possibility of ‘revisiting’ or
‘revising’ the work
3. Data extracts not always the same
The research community must foster and support a culture of transparency and honesty which promotes good practice, recognises relevant interests or conflicts and deals with these openly and explicitly. This applies across the whole range of research activity from study and experimental design, generating, analysing and recording (including archiving) data, sharing data and materials.
Good Research PracticeOrganisations should publish standard procedures for signing off and archiving laboratory records and notebooks.
All researchers should consider at the research proposal stage how they will manage and share the data they will generate.
We believe that a structured approach to data management results in better quality data that is ready to deposit for further sharing. Grant holders must formally deposit all data created or repurposed during the lifetime of a grant in a responsible data repository.
Can You Archive the requested Data?Data providers that provide datasets for
Secure Data Environments
Typically allow project extension for competition of analysis
NO Too LongNO permission for archiving
• NHS Digital require the data they provide to be destroyed when the Data SharingContract or Agreement has expired. These can be renewed annually, but require re-approval.
• PHE provide data for one-year periods. After a year, the researcher must request anextension. At the end of the project the raw data and any sub-datasets created duringthe analysis must be destroyed.
• ONS and UK Data Service provide access to control data only users who have beentrained and accredited and their data usage has been approved by the relevant DataAccess Committee. Access is provided for maximum 12 months but can be extended
Can You Archive the requested Data?Does the Data Access Agreement od Data Sharing Contract specify
what to do with the data when project is finished
NO
Data can be Archived
YES
Data can NOT be Archived
In House Archiving
Third Party Archiving
Delete the Raw Data
Permiss
ion
Archiving and Accessing Data In House
Only Admins Users
Complete Project
Investigate DAA
requirements
Who is the data
controller?
Notify data owner for this action
Archive store Data
Securely
Contact Admins for Permission
Admins provide access
to the UsersTime-frame for access
Archive store Data Securely
If you are Separate SDE Project
No destroy requirement
Archive Data
Access Archived Data
Upon Approval
No DAAOnly Admins
Complete Project or DAA
has expired
Secure Transfer Data to an
Archive
Data Destruction +
Proofs
Notify data owner for this
transferArchive store Data Securely
Contact Data Owner for Permission
Archive provide access
to the UsersTime-frame for
accessSecure Transfer
for modified data
Archive store Data Securely
Original Data NO AccessPre-Agreed
Use Encryption Other Data Sources
Archive Data
Access Archived Data
Upon ApprovalNew DAAUse EncryptionNO Access
Archiving and Accessing Data with Third Party
Archiving and Accessing Data with Third PartyChoosing third-party organisation that could securely archive these data is desirablefor the following reasons:
• Data Processor would not continue to hold on to these data themselves, even ifaccess to the original project team, or any analysts, was prohibited
• The access to this data could be monitored by a third independent organisationwithin a ‘circle of trust’
• The data to be deposited with another institution could benefit existing or futureresearch programmes. Validation of results or revision of analysis would bepossible
• Would encourage the generation of metadata about the data deposited. Thiswould make data like health records easier to use and more discoverable.
DOCUMENT YOUR DATA
• Study level Documentation• Data level Documentation• Metadata Creation
Document your Data: Study-LevelStudy-Level
Documentation
General Info
Project History
Aims
Objectives
Hypotheses
Data Collection Methods
Sampling Design
Sample structure
Geographic Coverage
Quality Assurance
Data Validation
Checking
Proofing
Monitoring Over-Time changes
Methodology Changes
Updates on the data files
Information on Data Confidentiality
Think about access
Publications
Document your Data: Data-Level
Qualitative -> Data List
Interview ID Age Gender Occupation Location Place of Interview
Date of Interview File name
Transcript
Recording
Quantitative
Variable Names Variable Labels Variable Description
Units of Measurements
Value Code Labels
Coding Explained
Codes for Missing Data
Derived or Constructed Variables
Document your Data: Metadata
Title DetailsGeneral Number:Title:Principal investigator(s):Data collector(s):Data Classification in CRUK:
Background Information
Data Generation Process / Reliability
Time coverage – Timeliness
Variables
Potential Problems
How to access
Research Use
Coverage, universe, methodologyTime Period:Country:Observation units:Kind of Data:Universe:Data Classification:Time Dimensions:Sampling procedures:Method of Data Collection:Number of Units:Frequency of Release:Data Updated:Last Update:Next Update:Latest Year Data: Data held from CRUK:
Thanks for your Attention
Yannis Kotrotsios, Data Access Officer, Cancer Research UKYannis.Kotrotsios@cancer.org.uk
top related