Data sets that are to be integrated should be compatible and comparable.
Define a data dictionary early in the project to efficiently describe and document your data.
The steps followed to create data should be thoroughly and accurately documented to allow others to evaluate and interpret your results.
Whether compiled or entered from paper records, new digital data should be double-checked for accuracy to avoid input errors.
Problematic data should be flagged to indicate known issues so potential users are aware the limitations of the data.
Understand the geospatial parameters (resolution, reprojection, datum, scale) of spatially-referenced data when integrating geospatial datasets from multiple sources.