You might want to use existing data for your research to do a meta-analysis, to re-evaluate findings, to perform new analyses on existing data, or to validate a model. This page explains how you can search for research data.
Use your network
Publishing datasets is not yet as common as publishing articles, and many researchers still keep their datasets to themselves. You might then want to ask people in your network if they have datasets you could use.
Search the literature
Many articles you read are likely based on data. If you find an article that is closely related to your research topic, see if it includes a reference to a dataset, or if it contains supplementary data files. If not, you could contact the author and request access to the data.
Data repositories contain datasets created and published by researchers or by organisations. The easiest way to look for repositories is to use Re3data. This directory lists both multidisciplinary repositories and disciplinary repositories. Multidisciplinary repositories contain datasets from various fields, and disciplinary repositories contain datasets from specific fields.
Although Re3data covers many repositories, it does not cover them all. You should always search for repositories within your discipline. You can also find disciplinary repositories with open data here.
Search indexes of datasets
Another quick way to look for data is to search indexes of datasets. Unlike repositories, indexes do not contain actual datasets but list their metadata, such as title, data, and creator.
- DataSearch lists datasets from repositories, together with figures, tables and other data from published articles. You can preview the data files, so you can easily see if a dataset is useful to you.
- DataCite’s Metadata Search lists datasets that have been given a DOI, a unique identifier for datasets and articles. The database covers datasets from several repositories.