VOCATIONAL TESTS AND ABILITIES 265 
(2) The manual factor is best measured by the more complex of the 
routine assembling tests. ‘The measures afforded by the simpler manual 
tests are largely specific in character. 
(c) The reliability of the tests——The reliability of the vatious measures 
employed in the research was investigated in detail. ‘The coefficients were 
found to be generally high. Where the manual tests were concerned, 
reliability was found to depend upon the number of repetitions of the 
operation constituting the measure of ability rather than on the length and 
complexity of the operation ; and was found to be independent of the stage 
of practice attained by the group measured. ‘The routine assembling tests 
were found to predict the ability to which a subject would attain, after 
practice, to about the same degree of accuracy as they measured his present 
ability (0*’7-0°9). 
(d) The transfer effects of practice and of training —An extensive investiga- 
tion into the effects of (i) practice, and (ii) training, was also carried out. 
It showed that the effects of uninstructed practice at any one of the routine 
assembling operations failed to transfer to any of the other operations, 
whereas a course of training, involving exercises based on one of the routine 
operations produced effects which transferred to each of the other operations. 
II. Tue Past YEAR. 
Work during the past year has progressed along the following lines : 
(a) Further statistical analysis of the data.—The saturation of the various 
groups of tests with their respective factors has now been determined as 
follows, from data obtained from sixty elementary schoolboys : 
Two non-manual mechanical aptitude tests v. the mechanical factor (m), 
080, 0:71 ; together, 0°83. 
Ditto v. the general factor, 0-39, 0°36; together, 0-40. 
Three mechanical assembling tests v. mechanical factor, 0°39, 0°63, 0°51 5 
together, 0°73. 
Ditto v. general factor, 0°13, 0°42, 0°23 ; together, 0°31. 
Five more complex routine assembling tests v. manual factor, 0° 56, 0°65, 
0°48, 0°37, 0°61 ; together, 0°80. 
Ditto v. general factor, 0°35, 0°14, 0°27, 0°25, 0°16 ; together, 0°34. 
Four less complex stripping tests v. manual factor, 0°26, 0:61, 0°32, 
0°33; together, 0-60. 
Ditto v. general factor, 0-23, 0°09, 0°30, 0°20; together, 0°32. 
Similar determinations, with very similar results, have been made from 
data obtained from thirty-six elementary school-girls. 
(b) Development of new tests ——It was decided last year to concentrate on 
methods of measuring the group factors which the data had disclosed. ‘To 
this end, the following new tests have been devised : 
(1) Non-manual tests of mechanical aptitude.—Diagram booklets have 
been prepared for use in conjunction with the ‘ models ’ type of mechanical 
test, so that the subject’s response may now be obtained in the ‘ selective ’ 
manner. 
Two new sets of easier models have been devised and constructed for use 
in the upper classes of elementary schools. ‘These also involve the use of 
a specially prepared booklet. 
(2) Tests of the manual factor.—Six new manual tests, of the routine type, 
involving assembling and stripping, have been constructed. These involve 
various methods of winding and unwinding string from nails and of thread- 
ing string through beads and through eyes screwed into a board. They 
L 
