[Meng Hu]

I don't understand your comment. I said that histogram displays much more information than S-W, and that if you would ask me which method I prefer, it's clearly histogram. By far. For example, only histogram can tell you if techniques like poisson regression should be conducted. And that is a common pattern in crime/delinquency data, where you usually have things like : 90% with zero occurrence, 5% with 1 occurrence, 3% with 2 occurrences, 2% with 3+ occurrences.

[Emil]

I know that you like histograms, but they are balling on subjective analysis. Eye-balling is not good at telling whether a distribution is very unlikely to be caused by chance (low p value).

Histograms are of course useful for deciding which analysis method to use in some cases (when N isn't small), as you say.

I didn't have time to update it apparently, I will do it now.

[Emil]

Ok, here's a new revision, #3. https://osf.io/p9d5z/

It has some language changes, histogram, Spearman's correlations.

The reason I stopped using Spearman's correlations is that Pearson's cors are not very sensitive to violation of normality. Spearman and Pearson cors are almost always the same.

[Meng Hu]

So you have made the changes concerning the citation of Malloy's data, added the histogram and Spearman's correlation. I have nothing more to add.

I approve.

---

A little word about referencing.

For example,

[5] Emil O. W. Kirkegaard. Predicting Immigrant IQ from their Countries of Origin, and Lynn's National IQs: A Case Study from Denmark. Mankind Quarterly, 2013.

In all your publications, you usually cite like this, "Emil" first and then "Kirkegaard". In other articles, the order would have been reversed. And in general, I think people would look for "Kirkegaard" if they want to look for your name. So, using "Emil" first does not really help the research. Not for me, at least.

[Emil]

The references come in the order they are cited in. The names come in their ordinary order which I prefer over surname-first approach.

When using APA-style, references are ordered by author surnames instead of order of citation. Most of my papers are written in LATEX with the Vancouver style (numbered), not APA. When it's APA, it's because it's handwritten or not primarily written by me (e.g. Dutton and Kirkegaard, Fuerst and Kirkegaard, Piffer and Kirkegaard papers).

[Meng Hu]

I have forgotten to mention something. Outliers may be a problem in small sample size. But Spearman's rho is robust against it. You can add this information.

[Emil]

Spearman's is not robust against outliers I think.

[Meng Hu]

I have seen it many times in books that Spearman's rank correlation is "robust" or "more robust" against outliers. Jensen also used rho several times to check for the influence of possible outliers when he performs MCV.

[Emil]

Since r_s is just correlation on rank-orders, it is not sensitive to any change in values that does not result in a rank-order change. In that sense, it is more robust than Pearson's. See e.g. https://en.wikipedia.org/wiki/Spearman%2...n_fig3.svg

[Emil]

The new immigration report has come out. Every year the Danish Statistics Agency releases a report on immigration in Denmark. This year they have included quite a lot of more useful data. As a new thing, it includes GPAs for some countries of origin, but only for a small number and only for second generation. I will update the paper with these new data and analyses of them.

http://www.dst.dk/da/Statistik/Publikati...?cid=19004