Data Cart

Description

QRELGB denotes whether the creators of the 1881 database altered the relationship to household head as originally transcribed.

The following outline of relationship classification in the 1881 British census details why altering the transcribed relationships was sometimes necessary.

During the transcription process the LDS made a number of revisions to the relationship to head of household field. These alterations impinge on the manner in which the relationship to head of household has been coded.

For the purposes of the LDS' data entry program every household had to have a head of household. If the original census enumerator's book (CEB) contained a household (usually defined as being between a pair of double slashes) whose head was not defined as such that word should be added after the relationship as enumerated. e.g., a family which contained a woman and two children, where the woman was enumerated as being a 'wife' would become the head, i.e. WIFE (HEAD). A problem concerning consistency is compounded by the diligence of some enumerators. Consider a family containing a husband (temporarily absent) who lives with his wife and 20-year old son. One enumerator may have considered the son the head of household and enumerated him as such; another might consider the mother the head and enumerate her as such.

Exactly how the transcribers would approach this problem is unclear. If it was necessary to establish a head they were told to use five pieces of information within the CEBs to assist them in their choices:

• schedule number change of address inhabited strokes household strokes family relationships

Note, however, it is impossible to check this process internally because schedule numbers as they appeared in the CEB were not included in the machine-readable version. Transcribers were instructed not to resolve unusual relationships, rather to give them as enumerated with clarifications written in single brackets. They were categorically instructed to leave the original as enumerated if there was any problem. There is evidence to suggest that this was not carried out consistently.

If any member of a household was enumerated before the head of household transcribers were instructed to leave it as the enumerator recorded it but adding a note saying that this had occurred.

If a household consisted of co-resident family groups and the CEBs gave correct relationships to the head of household (not head of family) the original was not to be altered, for example:

//	WILSON	GEORGE	HEAD
	WILSON	MARY	WIFE
	WILSON	JOHN	SON
/	JONES	ROBERT	SON IN LAW
	JONES	ALICE	DAUGHTER
	JONES	JAMES	GRANDSON

However, it is unclear precisely what would have happened if this family was enumerated as:

//	WILSON	GEORGE	HEAD
	WILSON	MARY	WIFE
	WILSON	JOHN	SON
/	JONES	ROBERT	SON IN LAW
	JONES	ALICE	WIFE
	JONES	JAMES	SON

What would happen? One of two possibilities could be expected. The first would have been to alter the relationships to those as stated in the first example; the second would be to make Robert Jones's relationship HEAD (SON IN LAW) and leave the other relationships as they are.

These problems are compounded when considering household 'inmates', i.e., lodgers and boarders.

Where distinct households contained only lodgers the first one listed, regardless of age or sex would have the relationship head added as the relationship to the head of household in brackets. If a lodger or boarder was identified as such in the relationship column, and it was clear to the transcriber that s/he resided within that household, the additional head should not have been added, even if there was a single stroke in the CEB. It is unclear on what grounds this rule was applied, though surname and occupation are likely to give clues.

When lodgers or boarders were living with their own families and were given relationships to suggest this by the enumerator, the transcribers were instructed to ensure that one of the group should be identified as the head, for example:

//	WILLIAMS	ROBERT	HEAD
	WILLIAMS	ELIZ.	WIFE
	WILLIAMS	MARY	DAUR
/	SMITH	HENRY	LODGER (HEAD)
	SMITH	FANNY	WIFE
	SMITH	DAVID	SON

When lodgers or boarders were living with their own families and were not given relationships to suggest this by the enumerator, the transcribers were instructed not to make one of them the head of household. Note that transcribers were told to ignore the occupations or names of people only to use the information given in the relationship to head of household column.

//	WILLIAMS	ROBERT	HEAD	SHOPKEEPER
	WILLIAMS	ELIZ.	WIFE	SHOPKEEPER'S WIFE
	WILLIAMS	MARY	DAUR	SCHOLAR
/	SMITH	HENRY	LODGER	LABOURER
	SMITH	FANNY	LODGER	LABOURER'S WIFE
	SMITH	DAVID	LODGER	SCHOLAR

When relationships within families were given incorrectly to the head of household, transcribers were instructed to add the correct relationship within brackets:

//	JONES	MARY	DAUR (HEAD)	18
	JONES	JOHN	SON (BROTHER)	17
	JONES	ALICE	DAUR (SISTER)	13
	JONES	GEORGE	SON (BROTHER)	11

It is perhaps interesting to note that the LDS generally choose the eldest person, regardless of gender as head, rather than the eldest male, the latter probably being the chosen practice of most enumerators. Also, for example:

//	GREEN	EDWARD	HEAD	55
	GREEN	PHOEBE	WIFE	51
	JAMES	ROBERT	SON IN LAW	25
	JAMES	SUSAN	WIFE (DAUR)	25
	GREEN	RICHARD	SON	20

The transcribers were instructed to transcribe the relationships as enumerated, however the evaluators were instructed not to alter obvious abbreviations. For example, if the CEB says 'Daughter' and it was transcribed as Daur, it will be found in the machine-readable version as DAUR. A single exception to this was the head of household: whatever was found in the CEBs the machine-readable version will contain HEAD. (It is on this basis that we believe that the contents of the HOUSESTD field were allocated to Self or otherwise.)

What these rules boil down to is the possibility that there are two possible relationships which can be used in the coding and reformatting program. Depending on the circumstances the different relationships will be used.

The basic principle on which relationships were coded are shown below in a revised version of the 'principles of classification' working document. Revised coding principles for the British relationship codes for the 1881 census, found in the variable GB51A412, GB81A413, and GB81B412.

There are three basic principles behind this coding scheme:

1) that the trailing digit of the code denotes the gender of the individual, as follows:

odd	male
even	female
zero	gender unknown

Thus a male head is 11, a female head 12 and a head of indeterminate gender 10. (The exception to this rule is those people whose relationship code ends in 199 which is a form used for an undetermined relation.)

2) The codes fall into a number of distinct ranges:

0-1999	residential kin
2000-2199	visitors
3000-3199	lodgers and their familial inmates
4000-4199	boarders and their familial inmates
5000-5199	institutional inmates and their familial inmates
6000-6199	servants and their familial inmates
9999	others

The residential kin codes are grouped reflecting the vertical and lateral relationships to the head of the household.

The codes for familial inmates and servants are applied on the same basis as the head's residential kin. For example:

Head	10
Wife [of head]	22
Son	31
Daughter	32
Lodger's Son	3032

There are two problems with the data as produced by the LDS regarding relationship information. These are allowed for with the third principle.

3) Wherever additional information has been added in the transcription phase an attempt is made to make it useable.

To account for these alterations and 'corrections' a second code has been added to assist in family and household analysis. The basic principle behind this code is to show the 'real' relationship to the head of the household. Thus the relationship 'Son (Head)' when found in the data is assumed to mean that this individual is the son of an absent head of household and for the purposes of the LDSs exercise this person has become the head of household. In this case the first code is 31 and the second code is 11 [male head].

This secondary information, in brackets, is, with two exceptions always coded in the second code position. The first exception refers to the possibility that the information in the brackets has the same meaning as that without the brackets or if the information without the brackets is not a relationship; the second concerns those that are only present in brackets, thus:

Wife (Head)	22	12
Son (in law)	31	35
Unm (Head)	10	-
Bro Wife (Sis in law)	156	-
(Wife)	22	-

A different principle is used to classify those who are related to inmates. Thus:

Boarder's son	4031	-
Boarder (son)	4000	31
Boarder son	4000	31

The last possibility has been altered to deal with problems which may occur when inmates (especially boarders) are also kin.

A single caveat must be made here. Based on the (potentially unwarranted) assumption that when visitors have a relationship they have that relationship to the head of household. Thus:

Visitor son	31	2000
Visitor (Son)	31	2000
Visitor's son	2031	-

Note that for analytical purposes, in the second group the second code can override the assumption that the person is the son to the head of household by making him a visitor. The third person can be ignored because its primary code is in the 2xxx group and there is no secondary code.

			note
Boarder and father in law	125	-	(1)
Boarder (father in law)	4000	125
Boarder father in law	4000	125	(2)
Boarder's father in law	4125	-
Father in law (boarder)	125	4000
Father in law (boarders)	125	4000	(3)
Father in law and boarder	125	-	(4)
Father in law boarder	125	4000	(5)

Similarly:

Wife (Lodger)	22	3000
Wife (Lodgers)	22	3000

Notes:
(1) Kin takes precedence
(2) Treated as in brackets
(3) Not 4125 as per "addition rules"
(4) As per the first example
(5) Kin takes precedence, but the program will decide about the boarding status.

Note: also where the code is in the 5000-7000 range and there is additional information on kinship the default value is used, so the addition rule can be implemented, thus:

table> Domestic servant6200- Domestic servant (son)600031 s

Further (minor) rules:

• If only information in brackets it is coded in the first position. If the string has 'no relationship' followed by brackets then 'no onship' is ignored. If the string repeats information then only one is coded. If brackets occur before other information then it is treated as after. If the string contains information on kin and on an adoptive onship, kin takes precedence... but in other cases where potentially multiple kin the 'nearest' to the head is coded.

Comparability

QRELGB is only available in Great Britain.

Universe

Availability

• Great Britain: 1851, 1881, 1881