Tables

1. Error statistics

The tables linked from this section provide some detailed statistics on errors on the commercial census sites, intended to supplement the analysis of indexing errors in the first edition of Census - The Family Historian's Guide.

Note: The statistics were compiled in July and August 2008 and record the state of the data at that time. You can expect that each of the data providers will by now have corrected most if not all of the individual errors listed. Also, they are based on a tiny sample of just two enumeration books, and represent the work, therefore, of just two transcribers for each data service. The tables should not be taken as represenative of the general quality of the transciptions, even of other enumeration books in the same census, on the sites in question.

Typical transcription errors

  1. Ditto as surname
  2. Willaims instead of Williams
  3. Smth instead of Smith
  4. Borwn instead of Brown
  5. Jhonson instead of Johnson
  6. Geroge instead of George

Error analysis for two complete Enumeration Books

  1. The 1861 census for Pevensey, Sussex (385 individuals), from a relatively clearly written enumeration book.
  2. The 1891 census for part of Islington, London (496 individuals), the original less clearly written and with some idiosyncrasies to trap the unwary transcriber.

2. User-submitted corrections to the 1901 Census Index

The original 1901 Census Index was created by QinetiQ. As batches of corrections were made to the index, lists of these were added to the web site. In 2005, the 1901 Census Index was purchased by Genes Reunited, which preserved these lists of corrections but did not add to them.

The corrections files have now been removed from the 1901censusonline web site, but DC Thomson Family History, who are now the owners of the 1901 Census Index, have kindly given us permission to host a copy of the data here as a resource for those wishing to study the types of error made in census transcriptions.

The data offered here consists only of the user-submitted corrections to the records for individuals. The original data also comprised a number of high-level changes made by QinetiQ - replacements of dittos, corrections to ages over 100, and some corrections to the names of enumeration districts. We have not included these as they are specific to the 1901 digitisation project and are not of any broader significance.

All rights in this material remain with DC Thomson Family History.