Fathering Girls

Well of course you already know my FTAnalyzer program analyses GEDCOM files. Adding a report showing children born to over 55yr old fathers would be trivial.

 

That would be great - and perhaps it would a way to get more people using the program?

 

Possibly. Although I'm fairly sure that knowing what percentage of children born to over 55yr old fathers is female isn't possibly the highest priority for research for most people? I suspect being able to click to search the census or to see who they have yet to enter to Lost Cousins is perhaps a bigger draw?

 

It would be interesting, but then there are a whole host of graphs and stats that we could come up with, but I was keeping that for future roll outs

 

I think it's impossible to say what will prove to be the biggest draw. What I can say is that different features will appeal to different people, so the more ways there are to 'market' FTA to members the greater the takeup will be.

An important factor in marketing is the degree of commitment required from the user. If we invite members to download FTA and run a Gedcom through it in order to find out how many of their relatives became fathers after 55 they're not committing to do very much themselves - the program does it all.

But if we talk about searching the census or entering information on LostCousins, that's going to put a lot of people off, because they know they'll have to invest time and effort to get results. We all know that it's less time and less effort than they think it is, but convincing them wouldn't be easy.

So why bother? Once they've downloaded and installed the program they can more easily be persuaded to use it in other ways - indeed, the curious ones won't need any persuading. The trick is to get them to download it in the first place.

 

Yes, I agree with what you're saying Peter. I've had a number of conversations with Alexander regarding "wouldn't it be nice if it could do...." We've spoken about like a time line, on a map showing where people moved from census to census for example.

I think the main thing though is to get a more complete solution. What it does now is very good, but there's a few things that Alexander wants to deliver. There's no point people downloading the software for the over 55 challenge if it doesn't do some of the searches for example.

Alexander needs help with the documentation, I've made a start but if I have to do it all it will take time.

Once Alexander is happy with it, and it has some documentation, then I'm sure he could add some marketing "hooks"

 

Having developed the census search I was thinking that adding an individual search to the individuals tab might be an idea. Have a radio button that selects B/M/D and the same drop down that selects site and a double click searches for B/M/Ds. Much more tricky than census searches of course as which collections to search is a whole lot less obvious.

Time lines and map plots are an interesting idea but I've yet to think of a good way of presenting that data that works for someone with 20,000 people in their tree as well as working for someone with 20 people in their tree. The basic issue is coming up with a display that makes sense. There is also the issue of how much map data you need to look up. Google and Bing both throttle the number of lookups per minute. To get all the locations in my own tree looked up would take around 7 hours at the rate that is permissible.

I can understand your point Peter about something simple that people don't have to do much to get a result. I would argue that there is a lot of that there already. The data errors report, the list of locations & occupations that shows you lists of people at that location or with that occupation. I just think the number of people who would want a report of the gender of children born to a man over 55 would be vanishingly small. Ok it is of perhaps passing interest but other than providing some stats to prove or disprove a theory its not something that is going to be used regularly is it? Whereas other reports such as a time line or a map of where ancestors moved to is more likely to be of lasting interest.

Ideally therefore we need a report that shows really useful information, that is useful to a wide audience, but requires no effort from the user!! That is the challenge!! A hook to entice people.

 

This particular feature is only going to be used once. That's why people who've installed the program are going to start looking at what else they can do with it.

The point about the over-55s research is that it's bang up to date, it's newsworthy, and it's different. Members who take part will feel that they're part of something - and they will be, even if the program shows that none of their relatives fathered children over the age of 55.

The research I reported in the newsletter only made it so far as New Scientist, but if the results are interesting enough our follow-up research could bring some welcome publicity (after all, there are probably more over-55 men having kids now than ever before). It's connection with family history may be fairly tenuous - though it's not wildly different from the 'fertility census' in 1911 - but realistically where else is the data for a survey like this going to come from?

I'm sure there are other analyses we could do, but we have to start somewhere, and the sooner we can get FTA in the hands of some typical LostCousins members the better - the people currently on this forum were not randomly selected.

Think of it as a way of beta-testing the program. Remember - I had 20 years experience in software publishing so I've got a fair idea of the pitfalls and opportunities.

 

Absolutely I do very much welcome your input, my experience is in design and implementation of ideas, the marketing side of things is a dark art to me. However I can see the point you raise is a perfectly valid one essentially strike whilst the iron is hot.

 

Version 1.5.7.0 now includes a child age profile mini report (it could do with being made to look nicer). This is available from the Reports Menu. I split the fathers and mothers ages into decade levels and split the children into genders, so you get the raw numbers of children born to teenagers, 20 somethings, 30 somethings etc.

 

Well done! Any chance of getting a split at 55 so that our results are directly comparable with the research?

 

Done. Split in blocks of 5 and I threw in a bar chart too!!

 

Very impressive! Now, how do I figure out which of my female relatives supposedly gave birth between the ages of 50 and 54?

 

The perennial problem with family history is that dates are a somewhat vague thing The best the program can do is work with what it has. So it may be that there is actually a vague birth date of child and/or a vague birth date of mother. eg: dates BET xxxx and yyyy or BEF zzzz. It is also possible that it is a mother who has a lot of children and its the youngest child which has an imprecise date.

eg: Mother born 1824 child born BEF 1875. The best the program can guess is that the mother was 51 at time of child's birth. Now its likely that the child was born before the mother was 51, however the data in the GEDCOM isn't as precise as that.

 

Version 1.5.7.2 now has a report on the reports menu showing older parents.

 

Very Nice reports Alexander

 

Thank you, very helpful - that's enabled me to pin down an error in my tree where the baptism date has been used as the birth date. It also prompted me to do some research in a branch that I haven't looked at for a long time, so good news all round.

 

Thanks Tim. I've added a response to your Locations query does this answer the questions?

Yes most of the time where the data looks odd it is because there is some older data untouched for ages which has been mis-entered eg: has a typo, or as you suggest a wrong event recorded etc. There may well of course be times when its the program that is wrong and it is typically these oddities that highlights issues.

You are starting to see one of the main points of the program to delve into the data and show up potential problems. The idea being that fixing these problems can prompt investigation and can unearth new facts in previously dormant areas of our trees.

It can also prompt greater standardisation in how things are recorded. Especially with locations where it is all too easy to be lazy and record places in a multitude of different formats. Now often this wouldn't matter but now we have the prospect of one click searching having the locations formatted in a standardised manner can mean easier searching. It can also prompt discoveries for instance seeing individuals from different branches living in the same area which may prompt investigation into finding a connection.

We all will have some errors and inconsistencies in our trees. A large part of the program is prompting us to fix these inconsistencies and hopefully notice some options that we previously discarded because the facts didn't fit.

 

Very nearly I have replied to it now.

 

I agree, but we have to be careful not to add information without any justification.

For example, the address given in the census get more detailed towards the end of the 19th century, especially in rural locations, so it is tempting to backfit the later information - perhaps a house number, or a street name, into the earlier data. However we can't be sure that they didn't move to a neighboroughing property (some of my relatives certainly did).