Will someone who knows about statistics please explain something to me....

[seeker]

..It's about this bee people have in their bonnets about a name in the top 100 being "too popular"

Right. My dd is called Grace. When she was born, 13 years ago, the name was so unusual and wacky that some friends even thought we had had some sort of religious conversion and we thought that she was an answer to prayer! (we had been together a long time before we got round to having a baby.)

Now it's in the top 10, I believe- or at least it was last year. But surely that doesn't meant that every 10th baby girl in Britain is called Grace does it? Because I only know 2 others, and they are all, coincidentally the same age as my dd. There are none in my ds's primary school (200 odd girls). I have very rarely heard another child being called Grace in the park or anywhere, and there is only very occasionally one in the births column of the paper.

So what does being in the top 10 mean exactly?

It will depend on how many births there are in a certain year and how many babies are named a certain name. This may vary from year to year.

There is a website in the US called How Many of Me to check how many people have a certain name.

No, it's not a percentage.

So there are about 600,000 babies born every year. About 300,000 of them are girls. Does that mean that there were 3,000 Graces born last year?

And there are about 17,000 primary schools. So 1 Grace in every 3.5ish Reception classes.

Think of a bar graph showing the frequency of all girls names. There will be 10 which are somewhat taller bars than the next 10 etc etc down to the tail created by those misguided parents who want something 'ewneek' and/or can't spell. There must be hundreds or thousands of possible names, so the percentage with the 10 commonest is unlikely to be very high.

The top 10 varies somewhat year on year, with names drifting in and out of popularity, so there may well be more infant Graces than teenage at the moment. One of the sunday mags had a graph this week which was vaguely interesting. Apparently names which are seen to be very faddish can die out and not regain popularity for 2 or 3 generations.

No, it doesn't work like that. Supposing 299,990 of the girls had a unique name, and 10 were called Grace. Grace would be in 'the top 10' and also 10 times more popular than any other name. But the chances of actually meeting one born in that year would still be small.

Sorry, that wasn't to Grimma, who is quite right.

Sorry, that wasn't to Grimma, who is quite right.

Firstly it depends where the data comes from - ie whether from (eg) the Times birth announcements, which rather restricts the socio-economic background of new Graces, or all registered UK births, which I suspect would have a lower percentage of Graces.

Secondly it depends whether just first names or first and middle names are included - the latter would almost certainly increase the perceived recent popularity of Grace (ditto Rose, May and so forth).

And finally - no, Grace being in the 'top ten' doesn't mean every tenth child is a Grace (or even a Something - Grace). Out of the thousands of different names given to children they have simply taken the ten most popular - each of which will still only represent a tiny percentage of the total.

As to why names suddenly, in the space of a few years, become incredibly popular, there is an amusing chapter in Steven Pinker's The Stuff of Thought about the sudden rise of Steve - particularly scientific Steves - in the post-war years. It is perhaps a lesson to anyone thinking that they have suddenly hit on a great new 'original' name. (And before I sound smug my elder daughter is named after her grandmother (and great-grandmother) - but her middle name is none other than Grace!

So it is not sensible at all to look at the top 10 names and say "well I won't call my child one of those because it's too popular?"

Then has blinding flash of understanding and re-appears. Is it because there is an infinite number of names? The every 10 girl being called Grace would only work if there were only 300,000 possible names?

There are many more boys with top 10 names than girls, plus the boys names change less year on year. If you call a boy a top 10 name then he is very unlikely to be the only one in his school/class/office etc. So I would say it is very sensible to avoid the top 10 if you don't want a too popular boys name.

It will depend. In the 1920s and 1930s a top ten name would have represented up to 5% of the babies born each year. I think nowadays, as people become much more open to choosing different names (and sons are not automatically called John Jr), the percentage and therefore the number of babies with a specific name will be lower. People are choosing more unusual/less common names, which means the number of babies with a specific name will be lower.

Its funny how you can get clusters of names - when DH worked for ICI (when there was an ICI ) it appeared that the hiring policy was that they must have at least 70% Daves. In my own company (software) there seemed to be a Mike quota for a few years.

In the US for example (where the statistics are available), there are

5,101,567 James's and 51,300 men/boys with the name James Smith!

Have a look at HowManyOfMe - very intersting!

(there is only one other of me!!)