It says: 09-02-2014, 04:32 AM (This post was last modified: 11-05-2014 01:05 PM by Emil.). How do you normally make the word files into pdfs then?

Hi Emil,

I finally got the time to look at these papers. Both are interesting and should be published. Here some comments and suggestions, first for the Dutton paper about PISA and WAIS-IV:

1. Perhaps I overlooked it, but I did not find any mention of the date of the WAIS-IV standardization in the United States. I know it was pretty recent, but people who want to compare the IQs will want to know whether there any Flynn effect adjusted is needed when comparing the countries.

2. On top of page 2 you mention happiness as one of the correlates of IQ. In my experience, there is no positive correlation between IQ and happiness at the country level when per capita GDP is controlled, and there seem to be similar observations at the individual level. In other words, if we want to make people happy we should give them money, but not try to make them brighter. The other listed correlates do have specific relationships with IQ although some (e.g., corruption) are related more closely to GDP than to IQ.

3. Why are you using only the matrices test of the WAIS? Why not all subtests that measure reasoning ability and have not been modified to an extent that makes them uncomparable?

4. In Tables 1 and 2 you list the age ranges. Is the gap between age 39 and 45 intended, or should the 35-39 range actually be 35-44?

5. One way of estimating relative Flynn effects is to compare the performance of different age groups in cross-sectional studies. In Table 1 I see that the Finnish IQs (with USA as reference) are pretty constant at about 103 or 104 at all ages, suggesting that Flynn effects in Finland were about as big as in the USA. The Scandinavian sample in Table 2 has the highest IQs in the older age groups. This would indicate that Scandinavians born around the middle of the 20th century were brighter than Americans, but that they have lost much of this advantage since then. Is this an observation worth mentioning, or are sample size so small that random noise is the more likely explanation?

6. Bottom of page 4: The Greenwich meridian is not so much for international time comparisons (only indirectly), but mainly the reference point for degrees of longitude.

7. First line on page 5: The formula would more simply be expressed as IQ = (CPS - 517) x 15/96 + 100.

8. On page 6 you average PISA and WAIS with weighting for sample size, which is much larger for PISA than for WAIS. Here you have to consider that there is not only random measurement error, which is reduced by increased sample size. You also have to expect systematic error by non-representativeness of samples. Therefore it may actually be better to average the two with equal weight given to each. Or better, keep them separate and see if you can explain differences by the fact that we are dealing her with different cohorts: 15-year-olds in PISA, and adults for WAIS. If Scandinavians but not Finns do worse on PISA than on WAIS, this may indicate recent anti-Flynn effects in Scandinavia but not Finland. Perhaps the inflection point where the Flynn effect changes from positive to negative has been reached earlier in Scandinavia than in Finland.