Documentation and data quality

2a - What metadata and documentation (for example the methodology of data collection and way of organising data) will accompany data?

(Pay attention to metadata in project versus when archiving)

Science Europe DMP Guidance - annotated

2a.1
Indicate which metadata will be provided to help others identify and discover the data.

Identification and discovery of the data is an issue both in the active phase (internal in the project) and when data are archived and published. This will be useful to keep in mind when planning which metadata to provide.

2a.2
Indicate which metadata standards (for example DDI, TEI, EML, MARC, CMDI) will be used.

2a.3
Use community metadata standards where these are in place.

Many researchers might not be familiar with the notion of metadata standards. One approach could be to look for an archive much used in the community when planning where to archive/publish the data, and see what metadata standard is applied.

2a.4
Indicate how the data will be organised during the project, mentioning for example conventions, version control, and folder structures. Consistent, well-ordered research data will be easier to find, understand, and re-use.

This is important to decide and apply before entering the active phase of the project (and can be adjusted), and the same organisation may or should be maintained when the data is archived.\

2a.5
Consider what other documentation is needed to enable re-use. This may include information on the methodology used to collect the data, analytical and procedural information, definitions of variables, units of measurement, and so on.

These considerations should start prior to the active phase of the research, to prepare for the information to be collected simultaneously with data collection. In addition relevant and updated information needs to be made available as part of the archived dataset (e.g. embedded in the metadata or in a ReadMe-file).

2a.6
Consider how this information will be captured and where it will be recorded (for example in a database with links to each item, a ‘readme’ text file, file headers, code books, or lab notebooks).

If data documentation is recorded in a non-digital format (e.g. during laboratory or field work), explain how (parts of) the information will be digitalized, e.g. in ReadMe-files. Information on data processing and analysis steps should be included in the data documentation and kept up-to-date. This may include software versions, dependencies and workflows. In addition to parameters and settings, it is advisable to document approaches and decisions taken in the analysis process.

For more advanced use, use of workflow systems, reproducible deployment and build environments should be considered.

Science Europe DMP Evaluation Rubric: sufficiently addressed

• Clearly outlines the metadata that will accompany the data, with reference to good practice in the scientific community (for example uses metadata standards where they exist).
• Clearly outlines the documentation needed to enable data re-use, stating where the information will be recorded (for example a database with links to each item, a ‘readme’ text file, file headers, code books, or lab notebooks).
• Indicates how the data will be organised during the project (for example naming conventions, version control strategy and folder structures).

Coverage in RDA Common Standard for maDMP

[Properties in dataset]

• #metadata_tree: nested, see below

[Properties in metadata]

• #properties-in-metadata: description, language, metadata_standard_id (identifier, type)
• #metadata_standard_id_tree: nested, see below

[Properties in metadata_standard_id]

• #properties-in-metadata_standard_id: identifer, type

Other DMP guidance

Guidance from NFR

[Documentation, metadata, and data quality]

• What metadata will be used to help others identify and discover the data?
• What metadata standards will be used? Examples could be DDI, TEI, MARC, CMDI.
• How will data be organized throughout the project? This can be version control, file structure, conventions for file naming, etc.

Horizon Europe DMP Template

[2.1 Making data findable, including provisions for metadata]

• Will data be identified by a persistent identifier?
• Will rich metadata be provided to allow discovery? What metadata will be created? What disciplinary or general standards will be followed? In case metadata standards do not exist in your discipline, please outline what type of metadata will be created and how

[2.2 Making data accessible - Metadata]

• Will documentation or reference about any software be needed to access or read the data be included? Will it be possible to include the relevant software (e.g. in open source code)?

[2.3 Making data interoperable]

• What data and metadata vocabularies, standards, formats or methodologies will you follow to make your data interoperable to allow data exchange and re-use within and across disciplines? Will you follow community-endorsed interoperability best practices? Which ones?
• In case it is unavoidable that you use uncommon or generate project specific ontologies or vocabularies, will you provide mappings to more commonly used ontologies? Will you openly publish the generated ontologies or vocabularies to allow reusing, refining or extending them?

[2.4 Increase data re-use]

• How will you provide documentation needed to validate data analysis and facilitate data re-use (e.g. readme files with information on methodology, codebooks, data cleaning, analyses, variable definitions, units of measurement, etc.)?

FAIRsFAIR FAIR-Aware Additional Guidance

• With respect to the metadata standards: Find a suitable community- endorsed metadata standard in metadata registries (e.g., RDA, DCC). Preferably use domain and discipline- specific repositories to deposit your data in when using community- endorsed metadata standards (use Re3data to find a suitable repository).
• Confer with your local research data management specialist to determine the best metadata standard to use when no community-endorsed metadata standards are developed in your domain.
• Use controlled vocabularies for your metadata (and other documentation) to create unambiguous, reusable, and machine-interpretable information. Choose your vocabulary based on community standards and information you have on the use and openness. When available in your domain, choose a repository to deposit your data in based on whether they support your preferred vocabulary (find a suitable repository on FAIRsharing).

Knowledge for support staff

• Identifying domain-specific archives. Institutional and generic archive properties. Trusted research archives.
• What lies in interoperability
• Metadata standards. Linked data. Vocabularies/onotologies. Data findability (incl. indexing, metasearch, research catalogues).
• Data organization. Version control.
• Data documentation and domain conventions.
• Specific software: domain-specific, ELN
• Useful resources, generic and domain-specific.

Knowledge for users

• Metadata standard mostly defined by archive, how to identify archive
• Data organization in active phase
• Data documentation in the active phase and for preservation

Existing sources that can be reused

NTNU - DMP guidance

What information is necessary for future users (including your future self) to find and understand the data? Tip: Read Making a Research Project Understandable: Guide for data documentation (Fuchs & Kuusniemi 2018).

Documentation might for instance include details on the methodology used, lab protocols, codebook, analytical and procedural information, definitions of variables, vocabularies, units of measurement, any assumptions made, and the format and file type of the data. Consider how you will capture this information and where it will be recorded (for instance in a ReadMe-file when you archive your data). Wherever possible you should identify and use existing community standards.

Metadata is often defined as “data about data”, and metadata standards are standardized ways of describing data. For more information about metadata and metadata standards, visit the website How to FAIR > Metadata. You could also see the RDA Metadata standards catalog to look for metadata standards for your field or data type.

A common metadata standard is Dublin Core, which is a list of 15 standardized elements describing a digital resource. Often, research data repositories (archives) will use some version of Dublin Core when describing datasets. An example is the NTNU institutional archive, NTNU Open Research Data.

Also indicate how the data will be organised during the project, mentioning for example conventions, version control, and folder structures. Consistent, well-ordered research data will be easier to find, understand, and re-use.

UiT - DMP guidance

How will the data be documented so that they are comprehensible and reusable for yourself and others also in the long term? (According to best practice, research data should be documented in a ReadMe file which explains column headings, abbreviations etc.)

What kind of metadata standard(s) will be used? (Both open and restricted data have to be provided with metadata according to section 4.6 in the UiT guidelines. Some academic fields have established metadata standards, whereas other fields do not. Examine best practice in your field. See this page for an overview of established metadata standards.)

What standards and methods will be used for data collection/generating?

What is the expected file size for the data?

SIKT DMP

Metadata is data about data, i.e. information that describes other information through a set of attributes. Note that using a standard structures information in machine-readable form, which increases the visibility and retrievability of your data in the future. The RDA (Research Data Alliance) has an overview of both general and subject-specific metadata standards that can be applied to research data

It can be helpful to be consistent in naming files and folders during the project, e.g. by using file and folder names that clearly describes the content, what version they are, and when they were created. This makes it easier to identify the file(s) containing the information you are looking for. It also helps you keep control over different versions of files, and it makes it easier for others to understand and navigate through your work files. You can find examples of guidelines and procedures for naming conventions at Cambridge.

Describe other measures and processes that are implemented to document collection methods and processes during the project. If you do not intend to use a metadata standard, or there are no relevant metadata standards tailored to your research, describe what other strategies (e.g. whether you will use free-text documents, code syntaxes, etc.) will be applied to document the data.

EasyDMP

Documentation and data quality – covers how the data will be described (ie what metadata standards are used to describe the data) and how you will measure and record the quality of your data.

Describe how you will organise and document your data – this should include what metadata standards you plan to use, what data will be covered and how the arrangement of your data will help others to find and use your data. You should also describe any addiDonal documentation (such as electronic logbooks and user-guides or examples of how to use the data), how it will be captured and where it will be stored.

SND - Checklist DMP

Will the project use any established terminologies, ontologies, standards, or similar, to describe and document the material? Which ones? If you create your own terminology, will it be mapped against established terminologies? Why is this important? Many scientific disciplines have established terminologies, ontologies, and vocabularies (e.g. MeSH, ISCED, AAT, and ELSST) that can be used to categorize and document data materials. There are various standards that are recommended to follow (e.g. ISO 8601 for date, time, and time intervals). The use of standards and established terminologies simplifies the communication between people who belong to the same scientific field but can also make it easier to find material in, for example, journals. Sometimes, it may be necessary to create project-specific terminology lists if the existing ones aren’t sufficient or suited for the project. If you create a project-specific list, you should do a mapping, where you show which terms that mean exactly or almost the same thing as in other lists, but also which terms that are unique for your list. Mapping terms against other terminologies will improve the findability of the data material.

DMP Tuuli

How will you document your data in order to make the data findable, accessible, interoperable and re-usable for you and others? What kind of metadata standards, README files or other documentation will you use to help others to understand and use your data? Data documentation enables data sets and files to be discovered, used and properly cited by other users (human or computer). Without sufficient documentation the data cannot be reused. Documentation includes essential information regarding the data, for example a) core metadata (for discovery and identification) where, when, why and how the data were collected as well as b) descriptive information how the data is interpreted correctly using metadata standards, vocabularies and e.g. readme-files.

Tips for best practices

• Describe all the types of documentation (README files, metadata standards, vocabularies etc.) you will provide to help secondary users to understand and reuse your data. Repositories often require the use of a specific metadata standard. Check whether a discipline-specific metadata schema or standard exists that can be adopted.
• Consider how the data will be organised during the project. Describe, for example, your file-naming conventions, version control and folder structure.
• Use research instruments, which create standardised metadata formats automatically.
• Identify the types of information that should be captured to enable other researchers to discover, access, interpret, use and cite your data. See for example Qvain requirements (https://www.fairdata.fi/en/user-guides/qvain-user-guide/#QvainDataset)

2b - What data quality control measures will be used?

Science Europe DMP Guidance - annotated

2b.1
Explain how the consistency and quality of data collection will be controlled and documented. This may include processes such as calibration, repeated samples or measurements, standardised data capture, data entry validation, peer review of data, or representation with controlled vocabularies.

In the planning phase, describe all steps and routines to be applied that ensures the consistency and quality of the data, e.g. to avoid errors in the data collection and in the capturing and recording of the data. In the active phase, adjust routines if this is considered necessary or beneficial to this end.

At a conceptual level, data quality approaches and standards are described in the report Towards a data quality framework for EOSC by the EOSC Task Force ‘FAIR metrics and Data Quality’.

Science Europe DMP Evaluation Rubric: sufficiently addressed

• Clearly describes the approach taken to ensure and document quality control in the collection of data during the lifetime of the project.

Coverage in RDA Common Standard for maDMP

[Properties in dataset]

• #dataset_quality_assurance_tree: “Data Quality Assurance” [string, free text]

Other DMP guidance

Guidance from NFR

[Documentation, metadata, and data quality]

• Is there other documentation necessary to facilitate reuse? It can be description of methodology, information about analysis and protocols, definitions of variables, electronic lab books, code books, readme.txt files, etc.
• How is reliability and quality of the data controlled and documented? This can include processes such as calibration of measuring instruments, repeated measurements and samples, standardized data capture, validation of data entry, peer review of data or use of controlled vocabulary.

Horizon Europe DMP Template

[2.1 Making data findable, including provisions for metadata]

• Will search keywords be provided in the metadata to optimize the possibility for discovery and then potential re-use?
• Will metadata be offered in such a way that it can be harvested and indexed?

[2.2. Making data accessible - Metadata]

• Will metadata be made openly available and licenced under a public domain dedication CC0, as per the Grant Agreement? If not, please clarify why. Will metadata contain information to enable the user to access the data?

[2.4 Increase data-reuse]

• Describe all relevant data quality assurance processes

FAIRsFAIR FAIR-Aware Additional Guidance

• not covered

Knowledge for support staff

• Researchers might need help to present this explicit (might be self-explanatory to a researcher that one always calibrate an instrument for instance?)

Knowledge for users

• Data quality measure examples

Existing sources that can be reused

NTNU - DMP Guidance

How will the consistency and quality of data collection be controlled and documented? This may include processes such as calibration, repeated samples or measurements, standardised data capture, data entry validation, peer review of data, or representation with controlled vocabularies.

UiT - DMP guidance

When will the data be collected/generated? Is there need for extra hardware or software? Is there need for special expertise for collecting/generating data?

SIKT DMP

Examples of measures to control and ensure data quality may be the use of standardized methods and protocols to capture observations; checking precision and calibration of instruments; quality-check equipment to be used for audio or video recording; make multiple measurements, observations or samples; use standardized interview protocols and/or computer-assisted software when conducting interviews; and using standardized and consistent procedures when digitizing or transcribing data.

EasyDMP

Documentation and data quality – covers how the data will be described (ie what metadata standards are used to describe the data) and how you will measure and record the quality of your data.

Describe how you will control the consistency and quality of your data – this should include for example, calibration processes or control samples, data entry validation, peer-review of data samples etc.

SND - Checklist DMP

[Will the project use any established terminologies, ontologies, standards, or similar, to describe and document the material? Which ones? If you create your own terminology, will it be mapped against established terminologies?] Why is this important?

Many scientific disciplines have established terminologies, ontologies, and vocabularies (e.g. MeSH, ISCED, AAT, and ELSST) that can be used to categorize and document data materials. There are various standards that are recommended to follow (e.g. ISO 8601 for date, time, and time intervals). The use of standards and established terminologies simplifies the communication between people who belong to the same scientific field but can also make it easier to find material in, for example, journals. Sometimes, it may be necessary to create project-specific terminology lists if the existing ones aren’t sufficient or suited for the project. If you create a project-specific list, you should do a mapping, where you show which terms that mean exactly or almost the same thing as in other lists, but also which terms that are unique for your list. Mapping terms against other terminologies will improve the findability of the data material.

DMP Tuuli

How will you document your data in order to make the data findable, accessible, interoperable and re-usable for you and others? What kind of metadata standards, README files or other documentation will you use to help others to understand and use your data?

Data documentation enables data sets and files to be discovered, used and properly cited by other users (human or computer). Without sufficient documentation the data cannot be reused.

Documentation includes essential information regarding the data, for example a) core metadata (for discovery and identification) where, when, why and how the data were collected as well as b) descriptive information how the data is interpreted correctly using metadata standards, vocabularies and e.g. readme-files.

Tips for best practices

• Describe all the types of documentation (README files, metadata standards, vocabularies etc.) you will provide to help secondary users to understand and reuse your data. Repositories often require the use of a specific metadata standard. Check whether a discipline-specific metadata schema or standard exists that can be adopted.
• Consider how the data will be organised during the project. Describe, for example, your file-naming conventions, version control and folder structure.
• Use research instruments, which create standardised metadata formats automatically.
• Identify the types of information that should be captured to enable other researchers to discover, access, interpret, use and cite your data. See for example: Qvain requirements

Leif Longva

Contributor

UiT The Arctic University of Norway (UiT)

Lisbeth Jahren

Contributor

Norwegian University of Science and Technology (NTNU)

Svein Høier

Contributor

Norwegian University of Science and Technology (NTNU)