Remarks. This is an excellent test. It involves no knowledge which may not be presupposed at the age in which it is given, and success therefore depends very little on experience. The worst that can be urged against it is that it may possibly be influenced to a certain extent by the amount of reading the subject has done. But this has not been demonstrated. At any rate, the test satisfies the most important requirement of a test of intelligence; namely, the percentage of successes increases rapidly and steadily from the lower to the higher levels of mental age.

This experiment can be regarded as a variation of the completion test. Binet tells us, in fact, that it was directly suggested by the experiment of Ebbinghaus. As will readily be observed, however, it differs to a certain extent from the Ebbinghaus completion test. Ebbinghaus omits parts of a sentence and requires the subject to supply the omissions. In this test we give all the parts and require the formation of a sentence by rearrangement. The two experiments are psychologically similar in that they require the subject to relate given fragments into a meaningful whole. Success depends upon the ability of intelligence to utilize hints, or clues, and this in turn depends on the logical integrity of the associative processes. All but the highest grade of the feeble-minded fail with this test.

This test is found in year XI of Binet’s 1908 series and in year XII of his 1911 revision. Goddard and Kuhlmann retain it in the original location. That it is better placed in year XII is indicated by all the available statistics with normal children, except those of Goddard. With this exception, the results of various investigators for year XII are in remarkably close agreement, as the following figures will show:—