382 SECTIONAL TRANSACTIONS.— J. 



Dr. M. L. FiCK. — Intelligence Tests Results of Coloured, Native {Zulu), 

 Indian, and Poor White Children and their Social and Educational 

 Implications. 



Friday, August 2. 



Mr. R. J. Bartlett. — The Memory Value of Colour in Advertising. 



Ten advertisements in colour and the same advertisements in Black and White 

 were shown in various pre-arranged order in automatic books to 5,000 visitors at the 

 Advertising Exhibition held in London in 1927. The principal problem attacked was 

 the relative value of colour and black and white when all other things are equal, and 

 the method adopted was to exhibit black and white and coloured advertisements in a 

 second book in identical position to those occupied by their equivalent coloured and 

 black and white in the first book. In this way 2,500 reports on colour were balanced 

 against 2,500 reports on equivalent black and white advertisements. Taking one 

 hundred returns as a basis, it was found that the ratio of returns from colour to returns 

 from black and white from 400 records from each of ten different sets of com- 

 binations of the advertisements varied from 54'3 : 45'7 to 49'0 : STO with a mean 

 value of 51-5 : 48'5. The difEerence though small is statistically significant, and the 

 figures are very similar to those obtained by Nixon in America by an entirely different 

 method. Small initial advantages are magnified greatly owing to the fact that so 

 little time is given b}' the average man to advertisements, but it seems unlikely that 

 the proved value of colour in advertising can be due to this small advantage alone. It 

 is more probable that the ' pulUng power of colour ' is founded on a combination of 

 this small initial advantage mth other factors, and that one of these is the relative 

 amount of coloured and black and white advertisements carried by the particular 

 medium. 



Accordingly, the various combinations of advertisements were arranged to attack 

 the problem of the effect of the ratio of coloured to black and white advertisements on 

 memory value. Combinations in which two, four, six and eight coloured advertise- 

 ments appeared in the centre of a book with the outer leaves carrying black and white 

 advertisements were balanced by similar advertisements with black and white central 

 blocks. 



The greatest difierence between colour and black and white was obtained from 

 single advertisements at the ends of blocks of eight of the contrasting advertisements. 



The returns were equal with end blocks of four advertisements and central blocks 

 of six, and in every other case colour had an advantage which increased as the number 

 of coloured advertisements diminished. When books of all black and white were 

 compared with books of aU colour we approximated most closely to the conditions of 

 Nixon's American experiment and obtained figures identical with his — 52-5 : 47-5. 



It follows that in the interest of the advertiser the coloured pages in a magazine 

 should be Hmited, and that above a certain maximum, which would appear to be 

 somewhere between one-quarter and one-third, colour ceases to be worth the extra 

 cost, and that when coloured advertisements outnumber the black and white ones, 

 it is more profitable to use black and white, and also that in the present state of the 

 hoardings the use of black and white should pay handsomely. 



Comparing the black and white returns from different combinations, it is found 

 that in every case the total return from the black and white is greater than from the 

 complete book of black and white, which would seem to indicate that the ' puUing 

 power of colour ' is not obtained at the expense of the black and white advertisements 

 accompanying it, but that the advertising value of the whole is increased by the 

 introduction of colour. 



The experiments were arranged to cancel all effects of individual differences between 

 the various advertisements, but it is possible by analysis to throw these individual 

 differences into rehef and then, among others, the following interesting results 

 emerge : — 



(a) Straight-forward reahstic art secured considerably more returns than 

 modernist art. The ratio being approximately two to one. 



(6) Advertisements of known articles secured much better returns than those of 

 unknown articles. The ratio again being about two to one. 



(c) It is possible for good black and white work to secure better results than 

 average colour work. 



