[LINK] merry christmas foreign language speaking fellow linkers

[Scott Howard]

On Sun, Dec 27, 2009 at 12:35 AM, Tom Koltai <tomk at unwired.com.au> wrote:

> Extract from:
> 1351.0.55.002 - Research Paper: Modelling languages other than English
> spoken in Australia from Census data, 2000-01
> <
> http://www.abs.gov.au/AUSSTATS/free.nsf/log?openagent&1351055002_2000-01.pdf&1351.0.55.002&Publication&3E2A4663F70CDF38CA256FD4007F2A0B&&2000-01&31.03.2005&Latest
> >
>
> in other words, the ABS themselves realise that 8% of some language
> respondents (in this instance Greek) gave indeterminate answers.
>

I believe you have this document reversed.

This is an attempt to model language data based on other data, as if the
language data was not available.

The 8% you have quoted is the number of people who DID claim to speak Greek
at home, but where the basic modeling based on heritage would suggest that
they did not (eg,  those of "Australian ancestries", who would not normally
be expected to speak Greek at home)

That isn't to say that these 8% of people gave indeterminate answers as
you've stated - it just means that a simplistic model of using heritage to
determine language spoken at home could be wrong in 8% of cases in this
specific example.

But both anecdotally and empirically, the ABS reading of language spoken at
> home is IMHO incorrect.
>

Language spoke at home was a direct question asked in all recent Census's.
Obviously there is always room for error in census questions - and in fact
you can find the ABS's own analysis of that error probability for this
specific question at http://tinyurl.com/y9u46ow

Whilst their numbers may not be perfectly accurate, I do not believe the ABS
could mis-represent the number of people who speak a language other than
English at home by several hundred percent as you are claiming.

  Scott.