Gedcom Census - 1911 (and others) - Suggestions

Post by **NickWalker** » 12 Jan 2009 12:37

I would like to add support to Gedcom Census for some of the other columns introduced in 1911 in England and Wales and perhaps for census from other countries too where columns are not currently supported. I'd also like to ensure that the earlier censuses are dealt with adequately. I'm looking for some suggestions, advice, discussion regarding the best way to achieve this.

At the moment I'm only really interested in information regarding individuals so columns such as whether the dwelling is inhabited and number of rooms occupied I would expect to be added to the source text by the user.

I am considering having each census year potentially offering different columns for the user to fill in.

Here are my current thoughts:

1) 1851 census onwards: Support the Whether Blind/Deaf/Dumb column - where could this be recorded?

2) 1891 census onwards: Whether employer/employed/neither - I would usually expect this to be typedinto the occupation field in Gedcom Census (e.g. as Joiner (Employed)). Anyone think I need to do more for this? e.g. put the Employed into the note field of the occupation?

3) 1901 census: 'Whether working at home'. If I provided a tick box for this GC could automatically fill in the place and address for the occupation event created.

4) 1911 census: There are columns for 'Personal Occupation', 'Industry/Service with which worker connected', 'Whether Employer, Working, Working on own account'. Again the simple solution would be just to amalgamate these in the occupation field. For example if the 3 columns were 'Copper Roller Engraver', 'Calico Printer', 'Working', I'd enter it as 'Copper Roller Engraver for Calico Printer Industry (Employed)'. Does anyone think GC should do something different? e.g. put the Industry and whether employed into the note field of the occupation?

5) 1911 census: 'Completed years present marriage has lasted'. I could have GC work out (or ask) which marriage this refers to and if no marriage date is present, estimate a marriage date based on this. Alternatively I can just leave it to the user to enter in the source text.

6) 1911: Total children born to present marriage: 3 columns: Born Alive, Still Living, Died. This is a tricky one. There is a 'Count of Children' field in the family record but this is meant to be the total born to that family ever whereas the census records number so far. There is an Individual Attribute that could be used for the mother called 'count of children' which also has a date which could be used to associate it with the census and also could be linked to the census source. The problem is that GEDCOM defines this as 'The known number of children of this individual from all marriages', whereas the census just lists those born to the current marriage. Options might be: (a) use the individual attribute 'number of children' to store the total number born and put in the local note for this attribute that this is just for the current marriage and also put in there the number living and number dead. (b) Create a custom attribute for this (and maybe another one for number living) (c) Add the totals to a note for the marriage (d) don't do anything and leave it the user to enter into the source text!

7) Nationality: Store this in the individual 'National Or Tribal Origin' attribute.

I'd appreciate all your thoughts on this and also any other columns that censuses from other countries have that GC could deal with.

Best wishes

Nick

ID:3345

RalfofAmber · Post by **RalfofAmber** » 12 Jan 2009 13:55

I like option 5 on the marriages.
Regarding child count, it might be easy to point out if the living count is less different to the number of children entered though the problem of children leaving home probably renders this problematic. It would need to look at how old the marriage was and I can feel this would turn into a lot of logic that would never hit 100% of cases

PatrickT · Post by **PatrickT** » 12 Jan 2009 18:13

(1) should probably just be put in the note to the source as there isn't a standard attribute that covers it and it is probably a relatively rare occurrence.

(2)I suspect different people would want this handled in different ways. I try (not very successfully) to keep to a relatively small number of standard occupation descriptions so would personally prefer to see it go into the note field of the occupation.

(3) Good idea

(4) See (2) above.

(5) A very useful addition. Would it also be possible to show what the figure 'should' be from existing data in the same way as for age and place of birth?

(6) I think this should be (c) or (d), preferably with an option to choose. Would it also be possible to highlight where the census data conflicts with existing data as in (5) above?

(7) Seems OK.

Thinking about US Federal censuses, the columns that could also benefit from similar treatment are:

1900: Month & Year of Birth; Years Married; Mother of how many children; Number of children living;Year of Immigration; Whether Naturalised

1910: as above except Month & Year of Birth

1920: Year of Immigration; Whether Naturalised; Year of Naturalisation

1930: Age at First Marriage; Year of Immigration; Whether Naturalised

All easier to suggest than implement I suspect!

JonAxtell · Post by **JonAxtell** » 12 Jan 2009 22:00

1) There is a physical description attribute (DSCR).

2) I would think keep it simple and just put the actual occupation in the occupation field. Extra information such as how and why should go in the notes. Something to think of when putting data into attributes/events is how programs uses them in generating reports.

3) Good idea.

4) Same as 2. KIS.

5) If GC can estimate the marriage date, then it can add the date (as CAL 9999) to the marriage event, creating it if necessary. It should also add a note to explain how the date was calculated. It should be a configurable option to allow this to happen, some users might not want it. You can see after feedback if the config stays in future versions.

6) I would think that just getting GC to check the number of children already in the family against these values and to indicate to the user if there is a discrepancy. The user to sort out the discrepancy themselves, or allowing the user to add a note or set a flag that future work is necessary on the family.

7) Yep, NATI is a good place for this. A bit more adventurous is to check the birth place against the nationality using the last field in the place (the part after the last comma). This would require the user to use a pretty standard form of documenting the country. So rather than just record a city & county and assume UK because the user is only doing UK families, users would need to specify the country for every place. This is good practise anyway as you shouldn't make assumptions when doing genealogy! In terms of configurability, the user could specify the terms they use for England, Wales, Scotland, etc.

tommy166 · Post by **tommy166** » 13 Jan 2009 11:28

My 5d worth

1) I just record this at present on source text. Maybe it could be added as personal note.

2) Probably best left to individual to record as they like. I agrree with Jon - KIS.

3) That sounds reasonable idea. I've not bothered with it as many addresses I have are just villages.

4) So far, I have done like Nick and incorporated any extra information in the occupation itself. I've also noticed some returns where the fields haven't been completed very helpfully or where Occupation info has just been repeated.

5) Sounds a good idea as long as it would attach the marriage date to the relevant couple (where either or both had several marriages). Obviously I think it would need to check if there were a date already recorded.

6) I think this would be a minefield as it's only a snapshot at that date. Maybe custom attribute(S) for the woman if there is enough support for it.

7) Sounds very logical

Just as an aside how are people recording references for the 1911 census given the different coding shown on the released images?

GladToBeGrey · Post by **GladToBeGrey** » 13 Jan 2009 14:55

1) 1851 census onwards: Support the Whether Blind/Deaf/Dumb column - suggest this should be recorded in DSCR as long as a Date and Source link to the census are also added (they may not have always been DDB); also appended to the AutoText (I add this manually now, if present)

2) 1891 census onwards: Whether employer/employed/neither - I don't type this into the Occupation field; at some stage I'd like FH to support management of lists of Occupations (as per Place now) - adding the Employment details to the Occupation would create multiple versions of each Occupation. The additional information should go in a local Note to the Occupation (this is where I put it now).

3) 1901 census: 'Whether working at home'. If I provided a tick box for this GC could automatically fill in the place and address for the occupation event created. Agreed. I do this now.

4) 1911 census: There are columns for 'Personal Occupation', 'Industry/Service with which worker connected', 'Whether Employer, Working, Working on own account'. Don't amalgamate them in Occupation, for the reasons given above. Put the Industry and whether employed into the note field of the occupation, as above.

5) 1911 census: 'Completed years present marriage has lasted'. Suggest the user should enter this as an accurate record of the census contents, even if they're wrong (a note can be appended if so). GC should not start 'inventing' the right data that 'should' have been in the census

6) 1911: Total children born to present marriage: 3 columns: Born Alive, Still Living, Died. This is a tricky one. There is a 'Count of Children' field in the family record but this is meant to be the total born to that family ever whereas the census records number so far. There is an Individual Attribute that could be used for the mother called 'count of children' which also has a date which could be used to associate it with the census and also could be linked to the census source. The problem is that GEDCOM defines this as 'The known number of children of this individual from all marriages', whereas the census just lists those born to the current marriage. Suggest b) or d) are the best options, for similar reasons ... don't 'misuse' existing Gedcom tags for another purpose, even if nominally similar to the standard.

7) Nationality: Store this in the individual 'National Or Tribal Origin' attribute. Agree

YouTube · Post by **Jane** » 13 Jan 2009 16:14

at some stage I'd like FH to support management of lists of Occupations (as per Place now)

Sorry Nick off topic, but you can already manage occupations with the Work with Data/Occupations tool as per places

GladToBeGrey · Post by **GladToBeGrey** » 13 Jan 2009 16:25

Not Nick, ... John [wink]

You're quite correct, I'd forgotten. I'd still prefer Occupation and Employment details recorded separately.

Just for info: the 1911 'Infirmity' column (Blind, Deaf, Dumb etc) is being redacted until 2012, as the Information Commissioner has ruled this to be sensitive.

Post by **NickWalker** » 13 Jan 2009 17:35

John:Jane was apologising for me for posting something that wasn't regarding the topic I had set (which of course I forgive!

)

John, I read with interest all your comments but didn't really understand this one:

5) 1911 census: 'Completed years present marriage has lasted'. Suggest the user should enter this as an accurate record of the census contents, even if they're wrong (a note can be appended if so). GC should not start 'inventing' the right data that 'should' have been in the census

Surely if someone is stating how long they have been married then this allows an estimated marriage date to be created if it doesn't already exist (with the census as the citation) in a similar way to an age being used to estimate a date of birth. Of course GC would have to check with the user that this is being added to the correct family record and would be optional. I'm not sure what you think GC would be 'inventing'.

I appreciate these comments from all who have responded, so please keep them coming. They are very well thought out and are generally in agreement so I'm already beginning to feel we can reach a consensus which will keep most people happy (and I can provide options where there are differing opinions).

Best wishes

Nick

jmurphy · Post by **jmurphy** » 14 Jan 2009 16:52

Nick -- just wanted to toss in the US perspective on these two points:

Nick Walker said:
5) 1911 census: 'Completed years present marriage has lasted'.

6) 1911: Total children born to present marriage: 3 columns: Born Alive, Still Living, Died.

Question 5 would be useful in the USA as well for 1910 (present marriage) and perhaps 1900 (number of years married, which I think most people would have taken to mean the present marriage).

Having GC calculate a marriage date would be very helpful (if it's not too much trouble) or the option to create the marriage event if none exists, even without the date.

Beware the 1930 (and 1940) Censuses, which ask instead 'AGE at first marriage' (often an important clue that the present marriage is not the first one).

For question 6, in 1900 and 1910 we get 'number of children born / living ' which presumably is the entire number of children born to that woman (not just the present marriage). 1940 apparently asks 'number of children born (not including stillbirths)'. I have been using the individual child count attribute to note the total children born and adding a note about how many are still living.

For my purposes, the columns which would be most useful to have would be the immigration and naturalization questions from the US Census.

Those are:
1930 Year of immigration / whether naturalized
1920 Year of immigration / whether naturalized / year of naturalization
1910 Year of immigration / (for males) whether naturalized or alien
1900 year of immigration / whether naturalized

(Whether naturalized was recorded as a one or two-letter code: [A or AL = Alien; NA = Naturalized; NR = Not Reported; PA = First Papers Filed])

The question then becomes whether GC should check for an immigration event and create one if one is not there.

(I have a custom attribute to record these codes. Since there is no standard FH place to record whether naturalized, I suppose it would not make sense to use GC to record the naturalization codes, unless one wanted to include it as part of the Auto-text feature.)

However, I can certainly understand how you might want to stick to the questions that all the censuses have in common, so if the question of immigration never comes up anywhere else besides the US Censuses, it wouldn't be worth it to add it in. (On the other hand, if that kind of question also shows up in Canada or Australia, putting it in might be worthwhile....)

Obviously you must do what is best for GC and FH users as a whole rather than cater to my laziness! [wink]

Thanks for your consideration.

Jan

RSellens · Post by **RSellens** » 14 Jan 2009 17:40

Hi Nick,
I would agree with most of the other comments.

1) use the physical description attribute, with approriate date/source
2) use the note field
3) yes
4) use the note field
5) do the same as the DoB to cal the date if no marriage present on the couple selected.
6) I like the idea of a custom attribute, but i also like one of the other suggestions of a cross check against the number of children the family and display a message if more than currently known.
7) yes

Richard

GladToBeGrey · Post by **GladToBeGrey** » 15 Jan 2009 11:45

Nick Walker said:didn't really understand this one:

5) 1911 census: 'Completed years present marriage has lasted'. Suggest the user should enter this as an accurate record of the census contents, even if they're wrong (a note can be appended if so). GC should not start 'inventing' the right data that 'should' have been in the census

Surely if someone is stating how long they have been married then this allows an estimated marriage date to be created if it doesn't already exist (with the census as the citation) in a similar way to an age being used to estimate a date of birth. Of course GC would have to check with the user that this is being added to the correct family record and would be optional. I'm not sure what you think GC would be 'inventing'.

Sorry - I misread/misunderstood your OP to suggest GC should insert the 'correct' figure for length of marriage if a Marriage date existed in FH and it disagreed with the census. [oops]

Thanks for pointing out my mistake; I support the suggestion to create a Marriage using the census information if no Marriage date already exists.

Post by **NickWalker** » 15 Jan 2009 12:26

Thanks John

treefrog · Post by **treefrog** » 16 Jan 2009 14:22

Who says the 'number of years married' question is necessarily going to be answered correctly anyway? I've got quite a few ancestors who might have had reason to pretend they'd been married longer than they actually had...

Post by **NickWalker** » 16 Jan 2009 15:04

Its probably about as accurate as the age given on the various censuses

jmurphy · Post by **jmurphy** » 16 Jan 2009 15:52

Let's say a couple is expecting and the marriage takes place after the baby is conceived. In later years, they mis-report their marriage by a year or more so that it will look as if they were married first and then the baby came along.

In that case the faux marriage year is still a clue that they were already in a relationship, and if you have no other information, it is good as a start.

I put years of birth generated from census ages in as 'calculated' at any rate, and the do same for marriage dates that are generated from age at first marriage or number of years married.

If CG users don't want to have the generated date, they can always say 'no'.

Jan

Family Historian User Group