232 PRESIDENTS ADDRESS SECTION J. 



by assigning numerical values to the several tests ; and the figure 

 known as the intelligence quotient alsoi allows of direct com- 

 parison between the mental levels of different age groups. But 

 it seems likely that eventually the Binet type of scale will be 

 abandoned in favour of a single test or group of tests applicable 

 to a considerable age range. This latter method of measuring 

 innate intelligence was first attempted by Burt, and an account 

 given in his " Experimental Tests of General Intelligenoe, " pub- 

 lished in 1909. Work of similar character has since been done by 

 Abelson, Wyatt, Bickersteth and others. 



Burt has given special attenticn to the selection of tests of the 

 higher mental processes which he has found to be more closely 

 related io general intelligence than those of the sensory and per- 

 ceptual levels. T"hus his method of using a groujj of tests appli- 

 cable to a series of age groups will enable comparison to be made 

 of the intelligence oif the different age groups in a way difficult 

 or impossible with the Binet scale. There is also the advantage 

 that the tests proposed are more useful than those of the Binet 

 scale for the selection of gifted children. The series of tests for 

 reasoning ability recently published is likely to prove of great 

 value for the selecticn of the abler members of a- school population. 

 Mention may be made of the absurdities tests suggested by 

 "Ballard, which, when extended by further examples, may also 

 be found most useful. 



No tests, either of scholastic or of general ability, can give 

 information of scientific value till it itself has been tested. A 

 test in order to be useful for the measurement of general mental 

 ability must fulfil at least the following conditions. It must 

 measure the ability which it is supposed to measure, and hence 

 should show a high degree of correspondence with trustworthy 

 empirical estimates of intelligence ; its results should vary only 

 with intelligence and not with age, sex, or social circumstances, 

 except in soi far as these vary with intelligence; it must be reliable 

 and consistent in the results which repeated applications give ; 

 it should be easily administered and easily marked, and hence a 

 group test is more useful than an individual test. 



So far, noi one test has proved completely satisfactory by itself 

 for the measurement of general mental ability, but when the re- 

 sults of a group' of tests are combined a high degree of precision 

 can be reached. 



The tests which have proved useful elsewhere will probably be 

 of equal value for the measurement of the mental ability of 

 Australian children. But accurate measurement is impossible 

 until we have re-standardized the tests in use elsewhere or have 

 invented and standardized our own. This is a most pressing 

 problem for the advance of our science. A beginning has been 

 made in the laboratory of the Teachers' College, Sydney, but 



