Data management A cornerstone for responsible conduct of research // .

Data management

A cornerstone for responsible conduct of researchhttp://www.ori.dhhs.gov/;

http://www.ori.dhhs.bof/documents/rcrintro.pdf

04/19/23 1RCR - data management

Core areas – Research integrity

• Data management• Animal subjects• Human subjects• Conflicts of interest• Peer review

• Collaboration• Publication/authorship• Mentorship• Research misconduct

04/19/23 RCR - data management 2

Outcomes

• Understand rules of data management relative to responsible conduct of research

• Understand roles and responsibilities of research staff

• Develop communication plan for dealing with data management issues

• Implement a system for responsible data management

04/19/23 3RCR - data management

DATA MANAGEMENT OVERSIGHT

Requires time and effort of the project PI (principle investigator)PI ensures that research team plans, implements and maintains data management policies and procedure

04/19/23 RCR - data management 4

DEFINITIONS/TERMS

04/19/23 RCR - data management 5

Key concepts

• Data Ownership: who has legal rights to data and who retains rights; can the PI transfer data to another institution

• Data Collection: reliable collection of project data in consistent, systematic manner + ongoing system for recording changes (validity)

• Data Storage: enough data should be stored to allow results to be reconstructed

• Data Protection: protection written and electronic data from physical damage; protection of data integrity, including tampering, theft

04/19/23 RCR - data management 6

Key concepts

• Data Retention: time for keeping project data + secure destruction of data

• Data Analysis: how raw data are chosen, evaluated, and interpreted for other researchers and the public

• Data Sharing: how data and results are disseminated to other researchers and the public

• Data Reporting: publication of conclusive findings (positive and negative) after project completion

04/19/23 RCR - data management 7

Defining data

• Any information, observations associated with a specific project, including experimental samples, technologies, and products

• Any factual information used as a basis for reasoning, discussion or calculation

• Examples: instrument scans, survey responses, measurements, observations, and data sources

04/19/23 RCR - data management 8

DATA OWNERSHIP

Who controls and has rights to the data.Who manages and uses the data.‘Ownership’ is complex, and involves the PI, the project team, the sponsoring agency, and research institution, and any participants

04/19/23 RCR - data management 9

Data ownership entities

• Sponsoring institution (university, research firm). Usually has ownership (employs the PI), controls funding and is responsible for ensuring the funded research is conducted ethically. PI has stewardship over project data, controls research directions, publications, copyrighting, patenting, subject to institutional review.

• Funding agency (NSF, NIH, …) federal agency, foundations, private industry. Agencies may influence publication and marketing of data.

• Principal investigator. Generator and steward of the project data, often retains rights to ownership. Industry usually owns the data. In academic situations, PI may be able to take research and data with them. There are data transfer policies.

• Research subjects. May have partial ownership of data, or some control over the research results.

04/19/23 RCR - data management 10

DATA COLLECTION

Data collection provides the information necessary to develop and justify research. A successful project collects reliable and valid data.

04/19/23 RCR - data management 11

Data Collection

• Data collection should:– Enable researchers to analyze

and assess their work,– Allow independent

researchers to replicate the process and evaluate results

– Enable the team members to do good data management

– Detail the rationale behind the project design

– Support decisions on expenditures and project directions

– Yield reliable data, valid results and test the hypothesis or questions

Includes recorded information, how it is recorded, and the research design.

04/19/23 RCR - data management 12

Data collection objectives

• Team members should know:– Research purpose– Methodologies chosen– Implementation of

methodologies– How the data were collected

and analyzed– What expected/unexpected

results occurred– What expected/unexpected

errors occurred– Results significance and future

directions

Data collection guidelines and methodologies are part of the research design. Data collection should be consistent and systematic.

04/19/23 RCR - data management 13

Collecting valid data

• Diligent record keeping is essential• Some projects keep written and electronic

records• Record keeping: bound notebooks (errors

marked, dated but not erased). Large projects require other methods

• Electronic records: security can be an issue• Policy/procedure: know the project’s design,

guidelines, standards for data

04/19/23 RCR - data management 14

How to keep records

• Notes: you must be able to reconstruct the work and justify your findings. What worked, didn’t work, observations, commentary. Communicate your notes.

• Personal notebooks: entries should be chronological, consistent. Start a new page for each day; no blank lines between entries

• Error notation: entries should be indelible. Mark and date changes.

• What to record: anything that seems relevant to the project, i.e., data/time, team members who did the work, materials, instruments, software, data and observations

• Transferring data: verify that data has been correctly entered, particularly in electronic media

04/19/23 RCR - data management 15

DATA STORAGE

Storing data safeguards your research and your research investment. You must be able to recreate findings, augment subsequent research, or establish a precedent.You must be able to reconstruct a project and its findings.

04/19/23 RCR - data management 16

Data storage

• Type/amount of data to retain: be able to reconstruct the project. Information should include raw data, statistics/analyses, notes, observations, products or specimens.

• Electronic data: thorough documentation to allow future use, easy storage format, back up system.

04/19/23 RCR - data management 17

DATA PROTECTION

The data storage plan should include security issues.Protection usually includes limiting access to the data.Electronic data storage requires additional safeguards.

04/19/23 RCR - data management 18

Data protection

• Limited access: who is authorized to access and manage stored data, questionnaire privacy, …

• Electronic data: – Access: IDs, passwords, centralized process, wireless

access?– System protection: anti-virus, up-to-date software,

firewall,– Data integrity: record original creation date, use

watermarking or encryption, back up files, ensure destruction when desired

04/19/23 RCR - data management 19

DATA RETENTION

Sponsor institutions and funding agencies often have specific requirements for how long data should be retained.PI decides when to end storage.

04/19/23 RCR - data management 20

Data storage

• How long should data be kept: DHHS requires that data be kept for 3 years after the end of the funding period.

• Continued storage: $ for storage vs. future need.

• Data destruction: shredding, electronic data – Erase, CyberScrub

04/19/23 RCR - data management 21

DATA ANALYSIS

Data analysis methods must be appropriate for the project’s needs and objectives.Team members should know the data analysis methods.

04/19/23 RCR - data management 22

Data analysis

• Methods: select methods appropriate to the research setting, type of research, and research objectives. Data analysis methods are usually an essential part of the project design.

• Team member responsibilities: all members should understand the data analysis plan and be able to interpret the results in the context of their study.

Converting raw data to meaningful content: choosing, evaluating and expressing your data.

04/19/23 RCR - data management 23

Data analysis design

• Analysis methods: accepted standards in field of study – data form, assumptions,…; link significance to causation (NIH),;

• Data use: – outliers, missing or incomplete

data sets, data alteration, data organization

– Forging (inventing), cooking (retain only those results that support the hypothesis), trimming (unreasonable smoothing)

– Amending: instrument malfunctions, changes in samples, deviations in the procedures

04/19/23 RCR - data management 24

DATA SHARING

Data sharing is the way in which research is accurately represented to the scientific community and the general public.Data sharing during the project should be done with care, as the interpretations may be impacted by later findings.Some sponsor institutions and some funding agencies have their own data sharing requirements.

04/19/23 RCR - data management 25

Data sharing and reporting

• data sharing prior to publication: implications of a data set may not be knowns; shared results might be used for individual gain; there may be immediate benefits

• Data sharing after publication: the data is now open; sharing of raw data is reviewable

• Obligation to report: some funding agencies have stipulations on reporting; Patriot Act, Freedom of Information Act

Data is shared to acknowledge a project’s implications, contribute to a field of study and stimulate new ideas.

NIH endorses sharing of final research data; expects and supports timely release and sharing of research data from NIH-sponsored studies for use by other researchers.

04/19/23 RCR - data management 26

TEAM RESPONSIBILITIES

Each member of a research team has a different role and responsibilities. These (roles and responsibilities) should be understood by all team members.Teams often include: PI (enables the project), director (controls the project), research associate (coordinates the project), research assistant (carries out the work), and support (statistician, …)

04/19/23 RCR - data management 27

04/19/23 RCR - data management 28

04/19/23 RCR - data management 29

04/19/23 RCR - data management 30

04/19/23 RCR - data management 31

04/19/23 RCR - data management 32

WHAT ARE DATA?

04/19/23 RCR - data management 33

DATA OWNERSHIP

04/19/23 RCR - data management 34

DATA COLLECTION

04/19/23 RCR - data management 35

COLLECTING VALID DATA

04/19/23 RCR - data management 36

DATA PROTECTION

04/19/23 RCR - data management 37

DATA SHARING

04/19/23 RCR - data management 38

RESEARCH TEAM RESPONSIBILITIES

04/19/23 RCR - data management 39