Statistical data referring to broader classes, groups or categories. Data is an average, a sum or some other derivation from the data on the level of an individual so that it is no longer possible to distinguish between the characteristics of individuals belonging to these classes, groups or categories (see example 1).

Examples: the number of unemployed people and their age in certain geographical regions or the statistical data about the occurrence of certain offences on a national level initially derived from statistical data of individual police counties.

Data from a research study in which an individual or individuals are prospectively placed in one or more interventions (that can include placebo or other form of control) on the basis of which the effects of these interventions on biomedical or behavioural results linked to health are evaluated.

Data collected or compounded from several, usually heterogenous sources that share one or more reference points, whereby at least one of the sources was initially intended for other purposes. This data is included into a new entity (see examples 2, 7).

Example: providing data on the number of universities in the last 150 years using different available sources (financial documents, official statistics, university registries), by combining data form a survey with data from official statistics about geographical areas (population density, number of doctors per capita, etc.), or using the RSS protocol for collecting blog posts or tweets.

Qualitative data (text, video, sound or images), initially created for other purposes, converted into quantitative data (expressed in matrices by units) by using coding techniques in accordance with previously specified categorizing schemes.

Example: coding manifests of political parties such as the study on the 2009 European Parliament elections (doi:10.4232/1.10204).

Data that is the result of an experimental research method with which you control some or all variables included in the hypotheses.

Genomic data refer to the genome and DNA of an individual organism. It is used in bioinformatics where they collect, store and process genomes of living creatures. Genomic data usually stems from sequencing techniques, but its meaning is broader than sequencing data. It can also include data that is not connected to the sequence, such as data from microarrays, real-time PCR panel data and pharmacogenomic studies (see example 3).

Geospatial data is any type of data with spatial coordinates that enable mapping to the Earth’s surface. It can represent physical objects, discrete areas or continuous areas. Discrete geospatial data is usually represented by vector data, made up of points, lines and polygons, whereas continuous geospatial data is usually displayed by raster data such as cell grids in which each cell has its value. Many applications in a wide range of areas create geospatial data (such as GIS, remote sensing equipment, GPS units, archaeological total stations, manual mapping and computer-aided design (CAD)) in various formats, such as images, vectors, text and tabular data.

Vector-based geospatial data includes tables listing archaeological sites along with their coordinates, text-based files (e.g., XML) containing coordinates and topology for historic road networks, electoral results for political parties by administrative area. Raster-based geospatial data includes satellite images, aerial photographs, scanned maps, and digital maps of elevations, vegetation, land use, sea surface temperatures, air pollution, soil types, etc.

Laboratory notebook is the primary research record. Researchers use laboratory notebooks to document their hypotheses, experiments and initial analyses or interpretations of these experiments. This designation is used both for traditional as well as electronic laboratory notebooks.

Data that is the result of evaluating specific properties (or characteristics) of beings, things, phenomena (and/or processes) by using previously established standards and/or specialized instruments or techniques (see example 8).

Data that is the result of research by observation, which includes collecting observations in real time (e.g., observing behaviour, events, development of medical condition or illness, etc.) without trying to control any of the independent variables.

Data recorded with mechanical or electronic devices in form that enables the data to be recalled and/or reproduced.

Example: images or sounds on a disc or magnetic tape (see example 4).

Data stemming from modelling or simulating processes, events or systems from the real world, often by using computer programs (see example 5).

Example: a program that models how household consumption changes based on indirect tax changes; or a set of data on hypothetical patients and their exposure to a medication, their initial medical condition and known unpleasant/harmful effects.

Data acquired through research in which systematic use of statistical methodology is applied on a selected population pattern to gather data, and to research and evaluate a specific characteristic of that population. This includes population censuses, sample surveys, databases from administrative registers and derived statistical activities, and questionnaires (see example 6).

Collection of research data is the result of research and is usually published as a collection in a national or an international trustworthy data archive (repository). It must be publicly accessible and documented in a form that enables repeating of published scientific findings. Its quality is evaluated on the basis of data, metadata and comprehensive support documentation (e.g., measuring instrument, research progress report, sample information, consent form, protocols, software; content can also be described in a ReadMe file when a repository does not have detailed metadata fields on its disposal). Its scientific relevance is demonstrated in its applicability for researching a broad set of applicative or theoretically meaningful problems (see example 6).

Corpus is a special collection of research data that is a uniform, standardly labelled and internally structured collection of authentic texts created in accordance with previously set standards and with a specific aim, available in electronic form and equipped with tools that enable a multifaceted searching and statistical data processing (see example 7).

Data set represents partial, incomplete research data published in a data archive (repository), thereby meeting the FAIR principles and enabling research to be repeated. The data is equipped with metadata, licence, permanent identifier and labelled in a way that makes its incompleteness evident to users (see example 8).