The data for January-September 1999 and January, March-May and July 2000, and March 2001 were revised to correct for two inconsistencies in the products.
1. The 1999 and 2000 files listed above were revised due to incorrect data in the variables PRLKDUR and PRUNEDUR for some cases. Individuals with values in the lower range of entries, usually 1-12 weeks, were most likely to be revised. We corrected both the DataFerrett datasets and the ftp files for these months.
2. Additionally, the March DataFerrett datasets for 1999, 2000, and 2001were corrected due to a corruption of the weights fields that caused incorrect tallies of data within the DataFerrett system. Only the DataFerrett data were corrupted, not the ftp files for this issue. Therefore, only the weights in the DataFerrett datasets were replaced. No revisions were made to the ftp files for this issue.