140 The Nation [Vol. 120,No. 3110 These figures, when taken with the first set, throw the whole matter of conclusions into a somewhat new light. The first observation, when looking only at the data arranged according to countries of origin of foreigners taking the tests, is that there is a progression from high standing in countries populated by the Teutonic race to low for those persons coming from the so-called "Alpine" and "Meditteranean" races. However, it will be noticed, as Dorothy Hallowell has said in an unpublished survey of all the data on racial tests to date, that "roughly it seems that the nationality groups having the highest mental ages are from the countries where the educational advantages are good, and vice versa, and the English-speaking countries highest of all." This conclusion is reinforced when we turn to the second set of figures, where the data are arranged according to the length of residence of the foreign born in this country. For with increased residence comes, apparently, increasing intelligence. There are two conclusions, which may be drawn from these results. One is assumed and maintained by Brigham, and is to the effect that since immigration has changed in the past twenty years from the North to the South of Europe, the difference in racial stock may account for the increased scores for persons of longer residence in this country. A more obvious interpretation, it would seem, would be simply that with greater understanding of the customs of the country, longer experience in dealing with typically American situations, and more knowledge of the language, all of which comes with longer residence, and only with this, the persons taking the tests were better able to cope with the problems presented and to meet the situations described in them. Indeed, Dr. Brigham himself may perhaps be coming to this view, since in a recent symposium on the tests he stated that what the tests show is not innate ability, but what he termed "scholaptitude" - the relation between a given test score and a criterion based on a school product. This opens up the whole question of how the tests are constructed, and how they work. They are empirically determined, usually by a trial-and-error process, and the questions are usually only inserted after the tentative test has been tried out on a large number of persons so that ambiguities, too large a number of Extremely difficult questions, and the like may be eliminated. What is finally presented to the subject is a standardized test, usually with the time element involved, which is scored in a rigid way so as to eliminate the personality of the score to the largest degree possible. The tests, however (and this has to do with the tests that involve language difficulties, although the criticism is applicable to a certain extent to those which are non-language), are printed and must be read. There are specified directions which must be announced by the tester. There are answers to be written or marks made according to printed directions. Now, it is not hard to see that the language difficulty is a real one for persons who are of foreign birth. This has been made the subject of a study by Margaret Mead, who conducted extensive researches on the children of a small New Jersey town, some of American parentage, and others of Italian, of comparable social background, during the last year. Her conclusions are well worth stating here in full: 1. The Italians are definitely inferior to the Americans if judged by test showing alone. Therefore, if grading and promotion were to be governed by test results the Italians would be placed clearly at a disadvantage in competing with the American children in these schools. 2. The scores of the Italians have been shown to be influenced by the factors of language, as demonstrated by the classification according to language spoken at home; by the length of time the parent has been in his country, this factor being somewhat interwoven with the language factor; and by the social status of the parent. 3. This study would indicate that Intelligence-test scores of foreign children, particularly group tests involving the use of language, are subject to vitiation by the above-mentioned factors. Classification of foreign children, in schools where they have to compete with American children, on the basis of group intelligence-test findings alone, is not a just evaluation of the child's innate capacity. Special attention is called to the second conclusion, especially the part which refers to the language spoken in the homes of the children tested. Miss Mead divides the Italian children studied into four groups, according to the language spoken mostly at home - those whose parents speak all Italian, those in whose homes Italian and some English are spoken, those who live where English and some Italian are spoken, and those whose parents speak only English. The results are striking - the figures given are for the averages of the groups arranged according to the "Index of Brightness": [chart] Language&Number&Index of Brightness Italian&82&65.00 Italian and some English&100&70.00 English and some Italian&64&73.90 English&23&72.93 [end chart] The importance of the language factor comes out clearly in this table, although the average for the last group loses its significance because of the small number of cases in it. It is even a question whether children who are bilingual are not at a disadvantage in the tests. A study made not so long ago by Saer in England on children speaking Welsh and English seems to show that the handicap is on the side of those who learn a second language before they know the first completely, when compared with others who have learned only one thoroughly. Thus, there is the difficulty of language in giving the tests. How, then, to test immigrants? Shall we give them tests in their own language? We already have literacy tests for immigrants - might it not be possible to give those wishing to settle here intelligence tests which would avoid the language factor and thus admit only those with the highest standings? There is another difficulty in the way, which is far from a minor one, and which is recognized by most psychologists. It is the difference in cultural background. Tests are made out of the experience and cultural background of the maker. And it is too often the case that this background is totally different from that of the persons it is desired to test. This, in Army Beta, or the non-language test, there were pictures in which something was to be checked as missing, or misplaced, or with which something else was wrong. One of these showed a tennis court, with two men playing, but without any net. Is it to be wondered that the Hungarian steel-worker from Pittsburgh or the phantom Negro from Alabama was completely at a loss when it came to checking the element missing? Such a test with the example, often seen, of a house in which the chimney runs only to the height of the roof and not above it, would certainly have been failed in by Plato and Aristotle, to put the case grotesquely; while, coming a bit