This is an esoteric one but it’s basically a method to try to make population sample datasets more representative of the population of interest than they would be without it.
Details #
I have been meaning to explore it forever as I first came across it in my work with the hugely impressive Dr. Blerta Bodinaku in Tirana and our late, much missed, colleague Dan Pokorny. Dan and Blerta had used it when she was collecting the data for our paper: Bodinaku, B., Evans, C., & Pokorny, D. (2024). Interpreting Differences in Questionnaire Scores in the Context of Cultural Location: A Country Case Study of Symptom Check List -90- Revised Data from Albania, Germany and the USA. Academic Journal of Interdisciplinary Studies, 13(4), 156. https://doi.org/10.36941/ajis-2024-0106, open access. The reference in the paper is a book and I am sure that Dan had read it, or the pertinent parts and I just found some access to the book at https://archive.org/details/excursionsinmode00tann/. That suggests to me that the Hamilton in question is one of the key “founding fathers” of the USA and the figure at the heart of the recently popular musical “Hamilton” (see https://en.wikipedia.org/wiki/Alexander_Hamilton) and that the method has its roots in the vexed (and now horrifically messed up) issue of representation in the USA senate relating to the numbers of electors in the states.
Enough of all that. What is the method? Well it’s fairly simple and the idea is sensible. In Tirana it worked to match the ratio of participants by binary gender, by three age cohorts (18-29, 3049, 50+ years) and by the 16 administrative units of Tirana so that recruitment stopped when the numbers were recruited into each of those 96 cells (2x3x16) that would be nearest to the proportions in the cells according to the most recent census of Tirana. The principle can be used to balance by other variables of course.
Try also #
Chapters #
Not covered in the OMbook but the issues about representativeness and not overvaluing the, inevitably, non-random “sample” datasets runs through the OMbook.
Online resources #
Can’t see any happening!
Dates #
First created 26.iii.25.
