To use the wordscores suite you will need a Stata version 7 or later.
The wordscores suite assumes that your texts are in plain (ASCII) text. If you are working with texts in a wordprocessor, you must first save each text in a plain text format.
If your texts are in portable document format (pdf), you can convert them using a commandline utility pdftotext. Windows users can find a copy of pdftotext at the wordscores homepage. Carsten Blum has made an Mac OSX version available. Linux users can install xpdf, which includes pdftotext, from their distribution cds, or find a recent version at Foolabs.Although there is nothing in the wordscoring algorithm that requires english language material, limitations in Stata's text processing capabilities may lead to odd-looking characters appearing when you use wordscores tools. This will not affect the results of any wordscoring analysis, and will probably be corrected in subsequent versions of Stata.
| Previous | Up to Table of Contents | Next |