3
$\begingroup$

I've got some data about people in Bristol. I have about 75,000 records. Part of my dataset is each respondants age.

Aside from this, I have also downloaded census data for Bristol, which gives me the number of each in each age category i.e. 25000 people aged 1, 26000 people aged 2.... 916 people aged 100 etc.

What I'd like to try and do now, is to test how similar my own dataset is, in terms of ages, to Bristol overall. Can I give a figure which represents how 'representative' of Bristol?

Could someone talk me through how to do this please?

$\endgroup$
3
$\begingroup$

If I was reading or reviewing your work, I would be more interested in knowing how your data was obtained (the data generating mechanism) rather than in comparing the age distribution between these two data sets. As a trivial example, if your data comes from maternity records then while the age distribution is likely to be similar, it is clear that your data is only representative of half of Bristol.

If you believe that you have a 'representative' data generating mechanism, and now you want to compare the age-distribution then you could simply count the number of patients in your data using the age categories provided by the census data. Plot these out, and compare visually first. If you need to perform a 'test', then a $\chi^2$ test comparing the observed to expected frequencies for the cells in the table would be a good starting point.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.