Goed Johan, jammer vir die kortaf kommentaar - dis my eerste op julle uitstaande werk. Wanneer jy absolute getalle in ruimtes (dan implisiet by die sentrum van die ruimte/area geplaas) soos distrikte of natuurlik by punte soos bv dorpe, wil plaas, gebruik jy puntsimbole (sirkels, vierkante, driehoeke, sfere ens). Die grootte van die puntsimbool varieer proporsioneel volgens die getal wat dit voorstel. Jy gebruik die choropleetkaart (arsering met patrone of kleure binne ruimte) om 'n verhoudingsklas uit te beeld. Die klas val dus tussen gekose of statisties-bepaalde grootte-waardes, maar is uiteraard nie absolute waardes nie. Hoop dit help? Hannes
Thanks for reading Johannes and for engaging. I generally agree with your sentiment that for absolute values you should use points with varying sizes (we actually do this in the second set of figures, but not the maps). If you look at any of my other papers/publications you will generally see this technique (e.g. my paper with Johan on the 1918 Influenza or my recent paper about the Marikana massacre).
In this case there is something else going on -- our figures and maps were explicitly designed to call reference to the visualisations of W.E.B Du Bois, one of the leading sociologists of race in the USA in the early 20th century. While he never made maps quite like thsese, in his visualisations choropleth maps are always the default, even when showing absolute values or proportional values (well, ranges of values, as we do in our maps). We developed our visualisations to be in conversation with his, and to pay homage to his pioneering work.
So, yes, I do agree with you that this is perhaps not the perfect choice (and perhaps in later versions we will change it), but there is some reason to our madness!
Goed Johan, jammer vir die kortaf kommentaar - dis my eerste op julle uitstaande werk. Wanneer jy absolute getalle in ruimtes (dan implisiet by die sentrum van die ruimte/area geplaas) soos distrikte of natuurlik by punte soos bv dorpe, wil plaas, gebruik jy puntsimbole (sirkels, vierkante, driehoeke, sfere ens). Die grootte van die puntsimbool varieer proporsioneel volgens die getal wat dit voorstel. Jy gebruik die choropleetkaart (arsering met patrone of kleure binne ruimte) om 'n verhoudingsklas uit te beeld. Die klas val dus tussen gekose of statisties-bepaalde grootte-waardes, maar is uiteraard nie absolute waardes nie. Hoop dit help? Hannes
Thanks for reading Johannes and for engaging. I generally agree with your sentiment that for absolute values you should use points with varying sizes (we actually do this in the second set of figures, but not the maps). If you look at any of my other papers/publications you will generally see this technique (e.g. my paper with Johan on the 1918 Influenza or my recent paper about the Marikana massacre).
In this case there is something else going on -- our figures and maps were explicitly designed to call reference to the visualisations of W.E.B Du Bois, one of the leading sociologists of race in the USA in the early 20th century. While he never made maps quite like thsese, in his visualisations choropleth maps are always the default, even when showing absolute values or proportional values (well, ranges of values, as we do in our maps). We developed our visualisations to be in conversation with his, and to pay homage to his pioneering work.
So, yes, I do agree with you that this is perhaps not the perfect choice (and perhaps in later versions we will change it), but there is some reason to our madness!
Poor cartographic techniques
You need to elaborate, Johannes