MICHIGAN ACADEMY OF SCIENCE. 65 



Results. 



Great individual differences were found in each test. The greatest vari- 

 ability was found for kinaesthetic imagery, the least for visual. None of the 

 curves of distribution are multi-modal. In each of the tests for concrete 

 imagery visual ranks first for the group, auditory second and kinaesthetic a 

 poor third. In verbal imagery inner-speech ranks higher than visual (verbal), 

 but when this inner-speech is split up into its two components, visual ranks 

 first, auditory second, and kinaesthetic third. The tests used probably tend 

 to over-emphasize the visual-verbal imagery. 



There is some correlation between tests A or B and C, and a fairly high 

 correlation between tests B and C. The person who ranks high in visual 

 imagery in test B will likely rank high in visual imagery in test C, even though 

 it is subordinate to inner-speech in the latter. There is a fair correlation 

 between clearness and dominance, although there are marked exceptions. 



No person was found who is limited to a single kind of imagery; nor was 

 any found to be wholly lacking in any one kind. Nor is there any evidence 

 of the existence of distinct combination-types, that is, of specific varieties of 

 "mixed" types. 



Both intra and inter-test correlations show clearly that there is a fairly 

 high positive correlation between the clearness of auditory and visual imagery. 

 Those persons who have little but visual imagery have a lower grade of visual 

 imagery than those with clear auditory imagery. 



Kinaesthetic imagery is shown to be a greater predisposing factor than 

 auditory toward the ure of inner-speech. In every case inner-speech corre- 

 lates higher with kinaesthetic than with auditory imagery. 



The correlations indicate the validity of the "U-L" test for the detection 

 of auditory-verbal imagery. 



University of Michigan. 



