This technical design document describes an approach to implementing support for the submission, review and approval of requests to obtain research data stored in Yoda.
- Research group A group of researchers belonging to a particular research lab or otherwise collaborating on a common research goal. E.g. the Brain Morphology Lab.
The supported workflows are described in this section. A sequence diagram mapping the workflow is shown below.
A researcher seeking to obtain research data stored in Yoda must be able to submit a request for this data.
A researcher may submit two kinds of data request:
- data requests made for the purpose of publishing the analysis results
- data requests for data assessment only (DAO; e.g. project-internal quality assurance)
In the first case, a research proposal must be submitted along with the data request. The research proposal must describe the purpose for which the researcher wants to obtain the requested data. No research proposal is submitted along with a DAO data request.
In the subsequent sections, the workflows for the processing of a data request for the purpose of publishing the analysis results will be described. DAO data requests are treated in a separate section.
A project manager of a research group to which a data request is submitted must be able to perform a cursory preliminary review of this data request, choosing whether the request is accepted for further review, rejected with the possibility of resubmission, or rejected with no possibility of resubmission.
Likewise, a data manager assigned as a custodian of data requested in a data request must be also able to submit a review of this data request, indicating as a recommendation to the project manager whether the data request should be accepted for further review, rejected with the possibility of resubmission, or rejected with no possibility of resubmission.
After a positive preliminary review and after review the data manager’s recommendation, a project manager must be able to assign a data request for review to one or more members of a Data Access Committee (DAC), or to reject the data request (either with or without the possibility of resubmission).
If a data request is assigned for review to the DAC, the project manager must be able to set the length of the review period.
The DAC members must be able to review research proposals assigned to them, but only within the review period set by the project manager.
A project manager must be able to give a final evaluation of a research proposal based on the reviews by the data manager and the DAC member(s), therein approving or rejecting the data request (either with or without the possibility of resubmission).
After final approval of a data request by the project manager, the research proposal of the data request must be preregistered by the researcher in the research group’s preregistry on the Open Science Framework.
The project manager must be able to confirm that the research proposal is properly preregistered in the research group’s preregistry on the Open Science Framework.
Once a data request and its accompanying research proposal have been approved, the data manager must be able to create a DTA specifying the terms and conditions under which the requested data is made available to the requesting researcher.
Once a DTA has been created, the requesting researcher must be able to view the DTA and register his agreement to the terms and conditions specified therein.
After the DTA has been signed by the requesting researcher, the data manager may release the requested data to him. The means by which the data is transferred to the researcher is outside of the scope of the data request module.
To keep the complexity of the module to a minimum, existing Yoda functionality and modules are used to implement the functional requirements.
To allow anyone, including newly registered Yoda users, to submit a data request, a public iRODS collection named “datarequests-research” must be created. When a data request is submitted through a web form by a researcher, a subcollection is created which will act as a folder for all files related to the research proposal. The data request and research proposal is then saved to this subcollection as a JSON-formatted file.
Because researchers only have ownership of subcollections they themselves have created, they cannot access data requests by other researchers.
Project managers and data managers must be able to view all submitted data requests and associated data (e.g. reviews). They are enabled to do so by membership of, respectively, the “datarequests-research-datamanagers” and “datarequests-research-project-managers” groups. Membership of these groups grant the appropriate permissions on the public collection mentioned above (these include read and write permissions). Because these permissions are recursive, they also apply to the subcollections made by researchers.
Upon assignment of a research proposal for evaluation to a DAC member, the DAC member is granted read and write permissions on the subcollection in which the research proposal is stored.
DAC membership is determined by membership of the “datarequests-research-data-access-committee” group, which acts as a pool from which the data manager may pick an appropriate member.
A DAC member can evaluate a research proposal through a web form. The evaluation is saved as a JSON-formatted file in the subcollection that also holds the research proposal itself.
Because multiple research groups may use a Yoda instance, multiple data manager and DAC groups may be created so that data requests meant for data of a particular research group can be handled by data managers associated with that research group.
A project manager can approve a research proposal through a web form that sets the value of the status AVU of the research proposal subcollection to “approved”.
After the data request has been approved, the data manager creates a DTA and adds it as a PDF file to the subcollection (through a web form). The researcher will receive a request to upload as signed copy of this document.
The DTA created by the data manager is made available to the researcher. Once the researcher has created a signed copy of the document, he can upload it through a web form. The signed copy is saved in the subcollection. The data may now be made available to the researcher by the data manager.
DAO requests follow the procedure of regular data requests, as described above, excluding the project manager’s preliminary review, the data manager review, the DAC review, and the research proposal preregistration. In other words, after submission, a DAO request is immediately offered to the project manager for final evaluation, after which the procedure is identical to regular data requests.