ABSTRACT

A US Small Business Administration (SBA) database on government guaranteed business loans always seems to have errors in names of cities. A cautionary tale for any journalist working with data involves the unintentional trap in a US government contract database in the years prior to 2001. The database contained items such as which agency awarded a contract, which company received it, the amount, the year, and the place where work on the contract would be carried out. The record layout acts as a road map to the database. The author also discusses the fact that people need the code sheet or codebook that goes with the database. It is not uncommon for journalists doing Computer-Assisted Reporting (CAR) to discover not only entry errors in the records but also problems with the record layout and the codebook. Journalists must pay attention to the format, US or European format, and make sure to import it correctly.