Raw data, unaccompanied by any documentation, is meaningless. When you find a relevant data file online, you'll want to also look for the documentation. Documentation is usually in the form of setup files, codebooks, data dictionaries, questionnaires, and/or guides to the data.
Setup files: these files allow you to read data into different statistical software programs
Data dictionaries: Similar to traditional dictionaries, data dictionaries include different variables and information about each variable. The name of the variable, definition of the variable, units of measurement, codes etc. can be included
Questionnaires: text of questions asked in a survey
Guides to the Data: Guides often include information about sampling methods, data collection, weights etc.
There are a variety of formal metadata standards which can be found in the Directory of Metadata Standards. Some of the more common metadata schema are listed below.
Metadata Schema | Discipline |
Data Document Initiative | Social and behavioral sciences |
Ecological Metadata Language | Ecology |
Qualitative Data Exchange Format (QuDEx) | Qualitative social sciences |
Darwin Core | Biodiversity |
Service Entry Resource Format | Earth Science |
NCAR README Guidelines | Earth Science |
Content Standard for Digital Geospatial Metadata | Geography |
Dublin Core | General |
Although metadata schema are very useful for comprehensive data description, learning and writing these schema can be time consuming. Instead, many researchers create a README.txt to describe their data. The README.txt file is a text file containing key information about the data, and often packaged with the data set. Some metadata elements that are often included in README.txt files are file names, methods for collecting data, dates of data collection, the title of any accompanying published articles, and information about the variables (such as definitions or units of measurement). There are a variety of other elements that you'll want to consider when creating your README.txt file so that you, your collaborators, andother researchers will be able to better understand your data. The following sources will help guide you in creating a clear and comprehensive README.txt file: