Population for surveys about surveys

[epiamsterdam]

interesting when the mapping people talk about this without actually ever doing it :) We have a standard population growth factor we would apply if census data is outdated.

[epiamsterdam]

We would usually get this from official census data from the government. Often times very outdated census data, so we would manually adjust the data to account for a population growth factor.
usually in a word/PDF document and then transferred into an excel sheet - but with differing formats.

[AmmerB]

@epiamsterdam within the mapping meeting / OSM team there were some discussion on standardising methodologies for this as well. (baseline data)

E.g. looking for data from some bureau of stats in government. If its outdated, how to you make accurate estimates. (old/other survey data estimates, estimate population growth, missing maps data). Might be interesting to see what practices are here.

[aspina7]

but doesnt msf do its own population count a lot of the time anyway @epiamsterdam ?
Anyway @dirkschumacher think it would be good to propose a standardised way of inputting this in to a formula for samp size calculation / sampling. Just in a dataframe format that is easily reproducible in excel (i.e. two columns one with unit name and the other with pop num)

[epiamsterdam]

In refugee camps and IDP settinsg we (MSF) would do the pop count. But that is usually the exception to the rule.
For sample size calculation the population size doesnt matter unless the population is <10,000 in which case you can consider to finite population corrections. Again, not very frequent that we would do this and impact on sample size and estimates are really minor.

[aspina7]

True keep forgetting is pop independent, ma bad.
Still useful to have template for merging with survey data In analysis with survey package