Binned age groups

[ben18785]

At the moment, we use the midpoint of an age bin as the reference for modelling. If bins are narrow, this is probably ok. But for wider age bins, we may want to consider incorporating uncertainty in the age of the participants.

[ekamau]

maybe naive question - does the binning or midpoint value apply for data with discrete ages? example - 1,2,3,4,5 etc

[ben18785]

I guess that the midpoint of a discrete age is just that discrete age? E.g. the midpoint of those aged 5 is 5

[ekamau]

agreed! just wanted to confirm ..:)

[yungwai-chan]

That is a good point @ben18785. I have analysed serosurvey using convenience sampling and have some doubts in whether the age of individuals within an age band is evenly distributed, which could affect the 'mean' age.

[ntorresd]

@ben18785 , it would be nice if we could have some discussion about this here under the light of what you've been doing with Junjie. Or should we close this one and update the discussion elsewhere?

[ben18785]

I think it's fine to keep this here, personally. I think Junjie should be able to introduce this functionality into the package fairly easily.