Scientific data are data that have been generated in the framework of scientific research projects. Due to increased automation of many steps in such research projects in a wide range of scientific disciplines, the amount of scientific data has increased sharply in recent years, and data sharing by way of databases connected to the internet has opened up new opportunities and challenges for the handling of scientific data. Beyond data generated within a research project, any kind of information — be it government census information, or pictures on Flickr, or baseball statistics, or logs of search engine results pages or clickstreams on the World Wide Web, or properties of encyclopedic articles such as this one — can in principle be investigated by a research project, and the term scientific data is sometimes also used in this context.
Scientific data are usually the result of testing a hypothesis or otherwise exploring some unknown territory at the borders of present scientific knowledge by means of the scientific method. In the context of prior knowledge or of the expectations generated by a hypothesis, a simple spreadsheet with numbers representing measured or calculated data may reveal a path towards curing a certain type of disease or a way to build vehicles with lower energy consumption or overall ecological footprint, or a hint on how to decipher an ancient language. Usually, the way from such an initial idea is long and has to be paved with considerable further amounts of data, some of which will point into new directions.