FLHIG
 

Family and Local History
Indexing Group

Selected articles from the Newsletter

This article first appeared in SIGGNL 17  pages 13 to 17 (February/March 1998)

The British 1881 Census Project:  Part I
Stephen C Young

[First part of a paper delivered at the Annual Conference of the Association for History and Computing, UK Branch on Wednesday, 13 April 1994 at the University of Hull: the other two parts were published in the following issues of SIGGnl. The paper was written and delivered before the LDS 1881 Census project was completed, but is published here because it is still of interest to genealogical indexers - Ed.]

It is with some anticipation that I present to you the details of the British 1881 Census Project. With my own graduate background in history and archives, it has been my belief and expectation since 1989 that the academic community, or more specifically those performing social history research not only here in Britain, but throughout the world, will greet the announcement of this database with a certain amount of amazement. Perhaps this is beyond the mark, but I do hope to convey to those present the sensational opportunities and potential soon to be available to those using this database. By explaining the methodology employed in creating this database I'm sure that each of you will be able to understand and consider its capabilities and limitations for research. Although the 1881 database was initiated for a purely genealogical or family history research function, I'm sure you will agree its aptitude for a range of other research is enormous, such as demography and prosopography, as well as studies in migration, occupation, family structure and mortality to name a few. First I will acquaint you with the background and intentions which initiated the indexing of the British 1881 census.

The Genealogical Society of Utah, which is supported and funded by the Family History Department of the Church of Jesus Christ of Latter-day Saints (commonly called Mormons) has been collecting and preserving government and local records for decades. In fact this year marks the centennial of the GSU's organization. This society has been on the forefront of microfilming technology since the 1940s and is currently applying the latest CD-ROM innovations in several immensely large databases available for public non-commercial use. Perhaps some or most of you are already familiar with and use the facilities of the eighty-seven LDS family history centres located throughout the British Isles.

Background

In 1987 Richard G. Scott, then Executive Director of the Family History Department, met in London with the British Genealogical Record Users Committee (BGRUC), an ad hoc group composed of organisations committed to providing access to and preservation of genealogical and historical records. An initial proposal was made to index the 1851 census on a co-operative basis but this was rejected primarily because so many local family history societies in the United Kingdom had already begun or completed and were selling index booklets for their local areas. Therefore a second proposal was made and accepted to index the 1881 census (one generation after the 1851). At the time it was the latest census released for public research [the 1891 census is now available]. The necessary licenses and agreements were procured from Her Majesty's Stationery Office and the Public Record Office. At first only England, Wales, the Channel Islands and Isle of Man were involved, but by 1990 the Scottish Association of Family History Societies successfully applied to have the 1881 census of Scotland added to the scope of the project. Once again the necessary licenses were negotiated, this time with Her Majesty's Stationery Office and the General Register Office for Scotland at Edinburgh.

The accurate transfer of the approximately 26,300,000 census entries for England and Wales, and the 3,700,000 entries in Scotland, from the original enumeration books to electronic media has been of prime importance. Actually the indexes being created by the British 1881 Census Project exceed the requirements of a simple index since they carry all the information originally recorded for each person enumerated in 1881, including address, full name, relationship to head of family, marital condition, age and sex, rank, profession or occupation, and place of birth. If details are recorded indicating a person to be deaf and dumb, blind, or mentally deficient, this is also included in the indexes. With each entry also appears the census piece, folio and page numbers where they appear in the original enumeration book, with the GSU microfilm number. As already stated, the GSU's sincere intent has been to successfully transfer the information from the original records into computers. A strict procedural strategy was therefore developed and implemented to this end, and it is the discipline applied which we will consider now.

Preparation

Through the willing instrumentality of the Federation of Family History Societies and the Scottish Association of Family History Societies, membership of the constituent local societies were enlisted to share in the responsibility of copying the census entries. Before the project went into the production state the GSU was busy preparing the paper copies of the census which were to be sent to participating transcribers. Each enumeration district was divided into a number of "batches". Often one parish could be a batch, whereas large cities may contain many batches because of their large population. This task was completed using a database created to assign a number to each batch of enumeration detailing its geographic location and piece number, and specifying the number of folios and pages for each. All this was done to ensure that every page was copied and accounted for before shipment to the U.K. Each batch was then checked for completeness before a "batch transmittal form" was appended and all was bundled up for shipment to England. These transmittal forms accompany each batch through all steps of production to ensure a record is kept of who and where every procedure was performed. All this may sound quite simple, but it required over a year to accomplish. The database is continually updated as new information is identified to further refine the geographic location of some batches. We estimate that approximately three tons of the photocopied census pages were shipped to England for distribution to the project co-ordinator of each participant family history society, and another four tons of blank transcription pages were likewise printed here and supplied to the societies. In addition the GSU published a booklet for distribution to all participants appropriately titled How to Transcribe the 1881 British Census: a companion volume was also issued for Scotland.

Transcription

The precision necessary for copying the information from the census is increased by the requirement that every page be transcribed twice, each by a different person. Both sets of transcription are then checked for any discrepancies by a third person and suitable corrections made. One set of corrected transcription and the census hard copy are then sent to the Project Management Centre at Garretts Green. The cardinal rule continuously stressed to the transcribers was "transcribe what you see, not what you think it should be". The driving philosophy is to duplicate the original census, including any historical errors. For example, if names are obviously spelled incorrectly, transcribers were instructed to copy that supposed error exactly.

Evaluation

When the transcription has been received by the Management Centre it is examined to ensure the transcription and hard copy agree and that vital detail is included on the batch transmittal forms for proper tracking of the batches. This is not a detailed inspection of all the transcribed entries but more a check to confirm such facts as geographical identification and whether any pages are missing, and to make such corrections where applicable. When the batch is in order it is shipped to one of the six Evaluation Centres, each staffed by LDS church members voluntarily assigned to this work. The very important process of evaluation ensures that the transcriptions are ready for data entry and that the person entering the data will not need to make any decisions about the accuracy of the information. Once again, this is not an effort to correct data, but only to check for missing entries or blank fields which have been overlooked in the transcription phase. When evaluation is concluded the batches are returned to the Management Centre for assignment to a data entry centre.

It should be noted here that facility has also been instituted to allow for entries, pages or entire batches too unclear to read on the photocopies or census films provided to the transcribers. LDS project volunteers examine difficult enumerations from the original records twice a week at the Public Record Office on Chancery Lane in London. This is another attempt to render the indexes as accurate and complete as possible.

Data entry

Precision in the transfer of data is once again enforced by the requirement of duplicate entry of each entry into the 167 project computers and over 200 home computers associated with this work. As with transcription, each entry is executed by different individuals. There are seventy data entry centres spread from Helston in Cornwall to Aberdeen in Scotland, each staffed by more LDS church volunteers. A software program designed by the GSU, called Universal Data Entry (UDE), is utilised for this dual entry system. Any discrepancies between both entries is instantly identifiable on screen and immediately corrected. Data entry personnel make no qualitative decisions about the data set before them regardless of whether it seems correct or not. When all pages have been entered, the computer diskettes containing the data are sent to the Management Centre prior to shipment to GSU headquarters in Salt Lake City. A backup of the data is kept at each data entry site in the form of a copied floppy disk and on the hard disk.

Tracking

A very sophisticated computer system has been developed to track the census information from the original enumeration pages to the final computerised format. Each batch is tracked by computers at the Management Centre in Garretts Green to ensure that the status and location of each batch is instantly available: whether in transcription, evaluation, data entry, or already in Salt Lake City. The reports which are required from evaluation and data entry centres are manually entered in the tracking system computers at the Management Centre. This is vital to know so that the final batches of any county can be quickly identified and accelerated through the system, thereby making that county available for the indexing process in Salt Lake. The original of problems or errors in any of the aforementioned processes can also be easily identified and corrected by keeping track of where the batch had been routed.

Ç TOP