Reply #8: Dunno.

[Igel]

I gave it a quick look. I don't trust the results.

1. Those with just primary education are shorter than those with no education or with secondary or greater education. This is an oddity crying out for an explanation. I didn't see it. I suspect that the poor in urban settings have a greater likelihood

2. They pooled all their data. This brings with it a variety of problems.

a. I didn't see where they adjusted for sample size, although they note that (as you'd expect) different ethnicities have different heights. If they sampled more Mayans in later years than they did in earlier years, it would skew the data.

b. Within a country there are differences in ethnicities, esp. in Africa--although Guatemala would show some variation. More European = taller. If they expanded the range of US aid programs from areas with greater mestizo populations to more indigenous populations, that, too, would show up as a decrease.

c. I didn't see where they controlled for urban/rural sample size.

Now, perhaps they did. These are obvious confounds and a good writer that wants to convince the reader will not only present all the evidence for his case but will also parry obvious criticisms. (An excellent writer will do this without the reader noticing it, for the most part.)

Then again, SES stands for a lot more than just income.