Incentives for Data Producers to Create ArchiveReady Data

1
Incentives for Data Producers to Create
Archive-Ready Data

• Margaret Hedstrom
• Jinfang Niu
• University of Michigan

NSF Award IIS-0456022
2
Why Do Producer Incentives Matter?

• Basis for cooperation between producers and
  archives
• Critical for efficient and affordable processing
  of data
• Important for quality of archived data
• The hidden knowledge problem

3
Where Are Incentives Operative?

• Data sharing policies that require researchers to
  deposit data from federally-funded research (NIH,
  NSF, NOAA, NASA, NIJ, etc.)
• Data Archive and Repository Requirements that set
  minimum requirements for data and documentation
• Transfers of records to archival repositories

4
What Types of Producer Contributions Are Expected?

• Data
• Complete
• Accurate
• Documentation
• Sampling Frame and Selection
• Data Collection Instruments
• Coding
• Data Manipulation
• Other Compliance Requirements
• Confidentiality, Privacy, Other Non-disclosure
  Requirements

5
Why is it Difficult for Producers to Comply?

• Requirements are not known
• Requirements are not explicit
• Extra Effort is Required
• Outside normal research/business processes
• Episodic
• Requires additional knowledge/skill
• Data Producers gain little (if anything) for
  their extra effort
• Weak Incentives and Weak Rewards

6
Producer Concerns

• Producers are rewarded for using the data
  directly, not for sharing it
• Producers lose control and exclusive use of data
  when they deposit it
• Benefits of depositing data are unknown or
  unclear
• Sharing data exposes flaws and errors
• Confidentiality may be compromised, especially
  via deductive disclosure when data sets are
  pooled.

7
Potential Remedies

• Make data deposit requirements explicit and known
• Provide sufficient support for compliance with
  deposit requirements (compensation for time and
  effort, tools, training)
• Make benefits of data deposit explicit and
  visible
• Notify producers when others use their data
• Revisit citation and attribution requirements
• Enforce explicit consequences for on-compliance

8
The Hidden Knowledge Problem

• Hidden Knowledge
• Difficult to articulate
• Taken for granted
• Not anticipated, overlooked, etc.
• Hidden knowledge is difficult to transfer from
  the original producer to an archives, but can be
  critical for reuse
• Standard practices rarely take hidden knowledge
  into account

9
Mitigation

• Only the data producer can reveal hidden
  knowledge
• Acknowledge hidden knowledge and its significance
• Encourage data producers and provide means and
  incentives to make hidden knowledge explicit
• Provide communication channels between producers
  and users