Glossary

API (Application Programming Interface)

A software interface that allows for computers or applications to talk to one another.

Badges to Acknowledge Open Practices

These badges are granted by the Center for Open Science (CoS) and mark an academic article as open. For more see https://osf.io/tvyxz/wiki/home/. There are three badges:

  • Open Data: Data deposited in a repository as public domain or with an open data license
  • Open Materials: Data (open or not) is deposited alongside digitally-shareable materials that explain the studies. These materials should be are publicly available and have enough detail and explanation that an independent researcher can understand the methodology that generated the data.
  • Preregistered: The study was preregistered in an institutional registration system. For studies that also included a pre-analysis plan there’s a Preregistered+Analysis Plan badge.

Code

High level programming files that can be used for analysis, data cleaning, data processing, or anything else that can be programmed on a computer. This is both a tool and potentially data. Code files may be included as supplementary material in datasets where appropriate.

Copyleft

A type of open license that requires that the original creator is cited and that any derived products maintain the same type of license as the dataset. Copyleft seeks to give credit to the creators while not limiting the rights of reusers, like a copyright might. The term is a play on copyright.

Conceptual Structure

The semantic meaning of a metadata element along with a controlled vocabulary. Used to increase interoperability.

Controlled Vocabulary

A controlled vocabulary is a list, information thesaurus, hierarchy, or another knowledge organization system (KOS) that is maintained by an organization or community of users. Typically they’re publicly available and have permanent URLs although there are exceptions like the APA’s Thesaurus of Psychological Index Terms.

Creative Commons (CC)

A nonprofit dedicated to open access to creative works (writing, images, music, etc). They maintain the CC copyright licenses (see: https://creativecommons.org/licenses/) and the CC0 public domain license.

Crossref

Crossref is an official digital object identifier (DOI) Registration Agency that assigns DOIs to academic publications.

Crosswalk

A crosswalk is a file that maps equivalences between the elements of two or more metadata schemas. This allows for the conversion of metadata from one schema to another, increased interoperability between datasets, and the enhancing of datasets by combining data from different fields.

Data Documentation Initiative (DDI)

The standard metadata schema for human survey data. The DDI Alliance maintains two metadata schemas in active use, the DDI-Codebook (DDI 2.5) and the DDI Lifecycle (DDI 3). They also maintain controlled vocabularies and other schemas. For more see: https://ddialliance.org/.

Data Information Knowledge Wisdom (DIKW) Pyramid

A loosely organized model that defines the relationships between data, information, knowledge, and wisdom. In this pyramid, data is a series of raw observations, stimuli, or symbols that are refined as information. Information are data that has been interrogated and made useful. It is processed data that has meaning and purpose. Next is knowledge. Knowledge is a synthesis of information over time from multiple sources combined with contextual information and experience. Finally comes wisdom which is implicit knowledge based on experience applying knowledge to different situations. Wisdom allows user to ask and answer why we would use information or knowledge in a certain way without apparent thought.

Data Uncertainty

The degree to which data is inaccurate, imprecise, untrusted, or unknown. All complex systems have a degree of uncertainty inherent within them. A dataset with a high proportion of NA values (missingness) has a high level of uncertainty.

DataCite

DataCite is a DOI Registration Agency that focuses on registering and managing DOIs for data and datasets. DataCite also maintains a citation standard.

Dataset

The dataset is the data that is to be preserved and shared. Each dataset has a designated community that defines the linguistic and conceptual needs of the metadata and documentation.

Derived Data Product

A derived data product is the analysis or aggregation of the dataset for reports, visualizations, or further analysis.

Designated Commmunity

A designated community is the imagined set of users of a dataset or archive.

DOI (Digital Object Identifier)

A DOI or digital object identifier is a persistent ID and URL for a digital object, whether an academic article, dataset, or derived data object.

Embargo

A period in which access to a work that has been submitted to a distributor or publisher is restricted.

Hierarchical data

Data organized in a hierarchy with parent and child elements. These elements are usually organized in a broader-than/narrower-than or a has-part/is-part-of relationship. This type of data is more flexible than a table of data since it isn’t limited to only two dimension.

Human Readable

Data or documentation designed to be read by humans rather than machines. Typically, this is in documentation/natural language form but there are exceptions where a file is designed to be human and machine readable, like YAML.

Information Thesaurus

A form of controlled vocabulary that maps semantic relationships between elements. These are typically organized hierarchically (broader-than/narrower-than) with synonyms (see also statements)

Internationalization

According to the World Wide Web Consortium (W3C), internationalization is taking steps to develop content or applications in a way that will work for users from any culture, region, or language. Data Team standards state that we use internationalized and multilingual controlled vocabularies where and when feasible in our reuse datasets.

JSON (JavaScript Object Notation)

A data serialization format used in APIs. It is derived from JSON but is language independent and typically viewed as lightweight compared to XML. It works by creating key-value pairs with records and subgroups surrounded by brackets.

Knowledge Organization System

A concept system or scheme used to organize materials (digital or physical) to retrieve or manage items in those materials. It can refer to a wide range of tools including subject headings, controlled vocabularies, topic maps, information thesauri, lists, ontologies, authority files, etc. Broadly speaking this is the material in the scheme rather than the structure or schema.

LaTeX

LaTeX is a document preparation markup language used in scientific documents and outputs a pdf. It’s a complex language with many packages and the ability to write macros, allowing for intricate formatting and making it the more flexible than markdown or HTML.

Linked data

Linked data is data organized as a triple (node-relationship-node or subject-predicate-object). This structure creates a graph and allows for non-hierarchical relationships between objects.

Machine Readable

A program or document designed to be read by machines rather than humans.

Make Data Count

An international project to standardize metrics on research data use, especially views, downloads and citations, by combining various community existing standards into an open framework. It relies on Crossref and DataCite’s Event APIs. See: https://makedatacount.org/

Markdown

A lightweight and simple markup language for creating formatted text. For a cheat sheet that covers most markdown syntax see: https://www.markdownguide.org/cheat-sheet/

Markup language

A set of rules that uses tags to define formatting and other elements within a document. It is both human and machine readable.

Metadata

Metadata is the context, definitions, descriptions of the values in a dataset. This can include labels, study information, authorship, titles, funders, etc.

Natural Language data

A file of plain text or audio that is designed for human consumption, following the rules of a grammar along with idioms and other linguistic elements.

Open License

The legal statement alongside a work (including a dataset) that allows free content and software to be distributed and used with few if any restrictions.

Open Science

Open science is the movement to make scientific research (including publications, data, physical samples, and software) and its dissemination acceptable to the public with limited to no restrictions.

ORCID (Open Researcher and Contributor ID)

A PID, in the form of an alphanumeric code, that identifies authors and contributors to scholarly works. ORCID also allows users to maintain a record of their work on their website.

PID (Persistent Identifier)

A unique and persistent identifier that identifies an agent, organization, location, or work. These are permanent identifiers, akin to an identification number from a government.

Preservation file formats

Preferred file formats to preserve data or datasets in perpetuity. These are usually lossless and non-proprietary. For tabular datasets, the preferred format is csv.

Repositories

A digital storage space for researchers and research organizations to deposit datasets and derived data objects associated with their research.

Social Science Research Data Repositories

Research Lifecycle Repositories

  • Dryad: An international open-access research data repository for complete, re-usable, open datasets. All datasets are licensed under a CC0 (Creative Commons Zero) waiver. There is a submission fee.
  • figshare: A repository designed for derived data objects and research outputs. figshare accepts data in any format making it more flexible than other repositories.
  • Open Science Framework (OSF): An open platform designed to manage data throughout its lifecycle and enable collaboration. OSF hosts a number of services including preprints, preregistrations, and meeting and conference support. It also offers useful integrations into other repositories and systems.
  • Zenodo: A general-purpose open-access repository operated by CERN. Allows for the deposit of research papers, datasets and research software.

Code Repositories

  • Github: Repository for software development and version control using git. There’s the ability to assign a DOI with Zenodo in Github. Owned by Microsoft.
  • Gitlab: A software development platform and version control system using git designed to be a set of collaboration tools and a code repository. Gitlab has built in continuous integration/continuous delivery (CI/CD) and DevOps features.
  • Bitbucket: A software development platform and version control system using git designed for teams. It is a product of Atlassian, the owners of Trello, Confluence, Jira, and Sourcetree.

Domain Specific Repositories

  • Humanitarian Data eXchange (HDX): A repository that hosts international humanitarian data. The repository is hosted by the Center for Humanitarian Data in the United Nations Office for the Coordination of Humanitarian Affairs (OCHA). While not necessarily a research data repository, HDX allows for metadata records.

Institutional Repositories

Schema

The structure of a metadata file, database or other digital object.

Tabular data

Data that exists in a table with an x and y position. Usually viewed as a spreadsheet.

Tidy data

A data organizing philosophy that organizes data as one row per object/observation with the properties as columns/variables. It is the cornerstone of the tidyverse R package. It’s an alternative name for a data matrix.

YAML (YAML Ain’t Markup Language)

A human-readable data serialization language that is usually used for configuration files and for data storage. It is an official subset of and compatible with JSON.

XML (eXtensible Markup Language)

The oldest markup and data serialization language used for representing structured information. XML typically has a schema and can represent data or metadata. For example, DDI codebooks are written in XML.