Gordon has encountered discrepancies with the results obtained from the
Advanced Person Search.........

There are 40 possible combinations of Gender, Marital Condition and
Relation to Head:

Male, Single, Head
Male, Married, Head
Male, Widowed Head
etc., etc.

In order to create a database of all Hutchinson entries in the index, I
queried each of these 40 possible combinations separately
(breaking out searches by age when the 300 limit was exceeded).
Theoretically, this should have provided the info required to
construct a database, which in addition to Name, Age, Birthplace, Parish,
County, occupation, etc. would also contain the Sex,
Marital Condition and Relation to Head for each of the 25,000 plus
Hutchinson's.

Using the excellent software developed by others on this list, I included
the Piece, Folio and Person ID references and should have
been able to sort the individuals into household groups and order them as
per the actual Census (by sorting on Person ID within
Piece).

The problem I ran into is that many of the same individuals appear in more
than one of the 40 searches. For example my query for the
Female, Widowed and Daughter combination for the surname HU*CH*S*N
produced a list of 1,735 individuals. A subsequent query on the
Female, Married, Daughter combination produced a list of 1,806
individuals. More than 1,500 of the individuals obtained by the
second query were the same individuals obtained by the first query (i.e.
all data identical including the Person ID). The query on
Female, Other, Daughter also produced 1500+ duplicates.

While I could delete duplicate entries (based on Person ID) I have no way
of knowing their true marital condition (per the index).

I would have thought that there would only be one Marital Condition
attached to each individual in the index. Is it possible that
when the Marital Condition column was left blank in the original census,
that these persons have been cross-indexed? I expect not as
this contradicts the info provided in the Help Screens. I thought the
blanks would only appear as "Other".

Gordon Hawthorne

P.S. The database does show some interesting anomalies (eg 46 Single
Wives - one of whom is male) but I do not know whether the
errors are in the transcriptions or the original census information).

______________________________