Say a country with sparse data reports the number of people who drank alcohol in the past month as a single data point, representing both sexes and ages 10-65. The goal is to be able to estimate alcohol prevalence for each sex separately, as well as more granular 10 year age groups (10-19, 20-29, etc.) based on that one aggregated data point and any other data. Is there a name for this process so I can look into it more?

  • $\begingroup$ This task is, at a suitable level of abstraction, remarkably similar to dasymetric mapping. $\endgroup$ – whuber Oct 8 '17 at 19:44
  • $\begingroup$ First of all, the "sparse" data you describe wrt alcohol consumption is, without question, available in a more disaggregate form from sources like the UN, WHO, Gapminder.com, Quandl.com, and others. In addition, population demographics by age and gender is available for every country in the world from these same sources. These disparate pieces of information can be put together and estimates made for any missing or blank cells in a saturated cross-classified table using an iterative proportional fitting algorithm. en.wikipedia.org/wiki/Iterative_proportional_fitting $\endgroup$ – Mike Hunter Oct 8 '17 at 19:51

Your Answer

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

Browse other questions tagged or ask your own question.