what are the principles of open data sharing

Question 1

what are the reasons to open data

Answer 1

• must be open if publicly funded
• facilitate data aggregation
• permit creative re-analysis
• assist with later scientific developments
• signals commitment to transparency
• public trust in scientists is heightened when data are openly available

Question 2

what are where should data be shared

Answer 2

• data underlying stats reported = personal, lab and project specific websites - data with personal identifiers - institutional repositories
• materials like behavioural tasks, test instruments, code books etc = open science repositories
• data repositories - supplemental materials attached to an article and stored on publishers website

Question 3

what are the FIR principles for data sharing

Answer 3

• findable
• accessible
• interoperable
• reusable

Question 4

what is databrary

Answer 4

• find clips
• share data
• archive data
• use data

Question 5

what is a DOI

Answer 5

• digital ID for intellectual property trading
• Digital data preservation (manage and retain data)
• digital curation (management and preservation of digital data over the long-term)

Question 6

what is GDPR

Answer 6

• European general data protection regulation
• safeguarding privacy of participants
• special category data
• personal data can be shared on the basis of public or legitimate interest on the basis of written consent from participant

Question 7

what is anonymised data

Answer 7

• it is not feasible to provide all raw data due to ethical, legal or other reasons - some data is likely to be better than none at all

Question 8

what are synthetic datasets

Answer 8

• mimic organic datasets by retaining statistical properties and relationships between variables
• no record in synthetic database represents a real individual

Question 9

what is participant consent

Answer 9

• level of data sharing (kinds of consent for certain information shared)
• confidential to research team but doesn’t promise to destroy data or keep it to the research team

Question 10

what is data sharing work flow

Answer 10

• curate data for sharing (sensible files/ labels)
• create metadata (README text file for data set)
• update/finalise data
  -update metadata
• submit data to repository

Question 10

what is README

Answer 10

• details of how study was run
• how analysis scripts are to be run
• list of folders & files
• study descriptions/link to pre-registration
• link to pre-print

Question 11

what is good about data sharing

Answer 11

• completeness (data and data descriptors supporting a studies findings)
• reusability - how readily available data can be accessed and understood by third parties

Question 12

what are the data sharing practices like in Psych

Answer 12

• 95% of papers didn’t contain identifiable open date
• data was held in image or pdf format
• comprised of tables but not raw
• data stored in journals or hyperlink without DOI
• often missing README, data descriptions/codebook
• exclusions of raw variables from which calculations were made

Question 13

what is the computational reproducibility in registered reports

Answer 13

• data from 118 from registered reports to independently computationally reproduce the main results
• 62 articles met inclusion criteria, 41 had data available and 37 had analysis scripts
• run 31 analyses
• only able to reproduce the main result for 21 articles

Question 14

what are the solutions

Answer 14

• develop standards
• educate about best practices that guarantee reproducibility
• overcome obstacles to sharing data
• fully describe the dataset
• to provide clear and practical open data guidelines
• ensure ling term accessible version of their data
• use third party repositories