FLHIG
 

Family and Local History
Indexing Group

Selected articles from the Newsletter

This article first appeared in SIGGNL 18  pages 5 to 9

The British 1881 Census Project:  Part II
Stephen C Young

[Continuation from Part I of the article in issue 17]

When the process of transcription, evaluation and data entry are finished and the computer diskettes are received at GSU headquarters, each undergo a computer check to ensure that the data is still readable after shipment. Usually the diskette is determined as undamaged after its long voyage across the ocean and a file folder, called a Tracking File, is created which includes the batch transmittal form and a truncation report printed from the actual diskette. Just as in England, careful records are kept in order to track each batch through the system in Salt Lake City, including the date it was originally received. Additionally, the identity of the civil parish is carefully established for each batch relative to the county in which it is located by using several resources. With this much accomplished, the diskette and folder are sent to Verification where the data-entered batch header (once again, geographic information) is checked for completeness and other data is examined. This includes automated inspections to determine if folio and page numbers are in correct sequence.

The next step is Audit , a random sample of batches wherein a detailed analysis of each entry is made. All corrections, as identified in the Truncation, Verify and Audit procedures, are made to the data. A goal of ninety-five percent or better accuracy was set and is maintained through these checks. It should be stressed here that no corrections are made to the actual census data. A backup is made of the batch on a diskette and the date the verification and audit was completed is entered into the Tracking File. With this much time, labour and resources expended, the data on these diskettes is obviously very valuable and steps are taken to protect it. Each week all newly-processed diskettes are sent to a Routing and Receiving System (RARS) where the data is copied onto magnetic tapes for storage. Once again, an inspection is made at this point to ensure that the data is readable and the diskettes are returned for continued processing. The data for RARS clearance is entered into the Tracking File and the Management Centre in England is notified, thereby allowing us to inform the relevant data entry centres to delete each of these same batches which have been preserved on duplicate diskettes and on their computer hard drives.

Now that the security of the data is preserved, a standardization team loads all the batches for one county from their diskettes into a Hewlett Packard (HP) mainframe computer in preparation for indexing. With the data loaded in the mainframe computer several reports are run to examine the field structure, ensuring it will standardize and sort properly when the indexes and lists are processed.

Indexing

I have tried to emphasise the gravity of the GSU s intent to successfully transfer the original information to computer. However, it would be misleading if I were to leave you with the impression that nothing has been changed. The publication of the 1881 census indexes on microfiche have forced some modifications to the data. Remember, the production of these indexes is the primary motive behind this project. Of course the several fields for each census entry are of standard size. Greater allowance is made for fields containing information of primary identification, like surname and forename. Other fields, such as relationship to head (4) and marital condition (4) are standardized into codes to save space. If a field size is not large enough to display all the original information, the data is truncated and a plus sign (+) is included to notify the reader of that fact. Additionally, /a/ two unique fields have been added to the indexes to assist the researcher further. A one character miscellaneous notes field may display an asterisk (*) denoting extra information from the original census entry not carried in the index entry. This could include a range of enumerator comments. A head of household field was also designed, programmed and included whereby each entry is tied to that information. For example, a head of household would show the term same in this field, while a wife, son, daughter, lodger, servant or whoever shows the full name of the head of that family. The only other fields standardized are the counties and census place parish, and this is allowed only because of these fields are utilized in the sorting procedures during indexing. The spelling of birthplaces is not standardized.

From the computerized data, the GSU creates three different indexes, a fourth recreation of the original census, the miscellaneous notes index and two separate lists identifying institutions and ships and vessels enumerated on special forms in 1881. The data is sorted by fields (surname, given name, age, birth place or census place). Computer printouts of the fiche indexes and lists are produced and examined to identify any image flaws or sorting errors. When the quality is approved the data is released for production. Two master computer tapes are created; one for storage and the other is delivered to GSU Information Systems where it is converted to a COM tape format. The COM tape is sent to our commercial vendor, Anacomp Inc., where the indexes and lists are produced as master tapes for use in printing copies of the indexes on microfiche. But before being released for production, the master tapes are examined for a last quality check. It should be noted here that Anacomp recently developed and is applying a new laser printing technology in the production of the 1881 census indexes, rendering the finished images even clearer and sharper than with previous GSU microfiche products. This technology is also currently being applied to the International Genealogical Index (IGI).

Lessons Learned

I was asked to comment on lessons learned from this project. Firstly, it is safe to say that there will never be another set of census indexes created as complete and complex as these, so this work is not only innovative, but will remain unique. A project of this magnitude could never be accomplished on a commercial basis and it is only through the dedicated efforts of so many volunteers committed to assisting the family history community in general that such a product can be realized. In an effort to assist the researcher, several details were added which have only slowed down and complicated the whole procedure. For example, the head of household field, the institutions and vessels/ships lists, and the miscellaneous notes index are really extra frills beyond what a basic index should contain.

Secondly, an important innovation applied in the Scottish census project is the fact that only one transcription is made which is copied twice, rather that two transcriptions checked once as in England and Wales. We have not found accuracy to suffer and the time consuming copy work is reduced accordingly. It should be noted, however, that the writing in the Scottish census seems to be much clearer or better preserved which has assisted significantly the progress of the project.

Thirdly, a national project of this scope has proved too labour intensive, time consuming and expensive. Future co-operative projects will be more localized and may involved collaboration between local LDS congregations and family history societies.

The Indexes

Each of the microfiche produced for this project has the capacity to display 460 index pages, with fifty-one entries per page, resulting in 23,460 individual entries per fiche. For easier identification of the different indexes, the eye readable portion at the top of each microfiche is printed in a specific colour. The marketing license for these indexes is held by the Federation of Family History Societies while the data content is copyrighted and "published by permission of the Controller of Her Britannic Majesty s Stationery Office". Each index page carries a similar message with the additional copyright of the microfiche indexes by the Church of Jesus Christ of Latter-day Saints. These indexes will only be sold to research organizations (family history societies, universities, libraries) and not to individuals.

Surname Index (Pink)

As its title indicates, this index provides an alphabetical listing of all surnames within each county. Its greatest facility is in the rapid identification of specific individuals, each being sorted first by surname, then forename (given name) and age in descending order. Therefore those persons possessing a common name, such as the inevitable John Smith, will all appear together within the county index with the oldest being listed first, followed by all others bearing the identical surname and forename in sequence until the youngest is identified.

Birthplace Index (Green)

Like the surname index, this research tool is sorted alphabetically by surname. However, it is here that the similarity ends. Instead of the entries being grouped together by forename, each now appears in an alphabetical sequence of birthplaces. It is this feature which provides the ability to gather together all those persons of a common surname who were born in the same locality but who have scattered to different parts of the county. This function is especially useful in reconstructing family groups which are particularly difficult to trace due to migration. Within this arrangement each person is then alphabetically arranged by forename and the by age in descending order.

Census Place Index (Orange)

This index functions in a similar manner as the birthplace index. Individuals are sorted alphabetically by surname and then by the census place in which they were enumerated. Hence, all those persons with the same surname who were living in a specific locality will appear together. Then each person is again sorted in sequence by forename and age in descending order.

Arranged as Enumerated (Yellow)

The participants in this project decided in the initial phase of planning that a representation of the original census would be valuable to the researcher. Since most family history societies in the U.K. are years away from being able to afford compact disc technology and have limited access to microfilm of the 1881 census, this product was part of the original agreement. If the Church was doing this project independently there would be no need to create this.

Miscellaneous Notes (Brown)

In each of the preceding four indexes the placement of an asterisk [*] in the note field signals the fact that supplemental information for an entry is available. It is simply arranged by surname, forename, age in descending order, and then by census place. Although most notes are contained on one line, extended information will be carried on successive lines. Such miscellaneous notes may contain information such as, the person listed was blind, deaf, or dumb, an invalid, or even additional facts about their occupation. It is a very varied listing of interesting and potentially important family history detail.

Editor's note

[Regarding the Birthplace Index, to take up a point made in my editorial, I have the minor quibble that it is not sorted by birthplace, but then I am carrying out a one-place study of Chesham Bois, Bucks, and it would be useful to have entries sorted by place of birth first, rather than name first., in order to find strays more easily The project was, I presume, directed to the needs of family historians rather than local historians and one-place researchers. To be concluded in the next issue - Ed.]

Ç TOP