COMPOSITIOX AXD PALATABILITY OF ICE CREAM. 6 
METHOD OF COMPARING THE DIFFERENT ICE CREAMS. 
The difference obtained in the ice creams made in these experiments 
was measured by means of a market where about 50 daily purchasers 
had an opportunity to choose from three different kinds of ice cream. 
Their choice was then taken to represent the degree of desirability of 
each kind offered. A nominal flat charge was made for each daily 
portion of ice cream. No attempt was made to charge according to 
a sliding scale based on the composition of the ice cream, because this 
would disclose the difference in quality and defeat the main purpose 
of the test. 
This is a new method of measuring the desirability of ice cream, 
and we believe it has many advantages over the score-card method. 
Fig. 2.— Three different kinds of ice cream were sold, with an opportunity to choose the kind preferred 
on the following day or days. 
Since it is not difficult to obtain the opinion of a much larger number of 
people with this method, it is probably more reliable than the score- 
card method. Instead of submitting a sample of each different kind 
or batch of ice cream to be scored with a score card, we first classified 
the experiments relative to the ingredients and limited the number of 
ice creams in each experiment to three, so that the major differences 
would be in the same general direction in each group of experiments 
Then while the treatment given the three ice creams and the condi- 
tions of freezing and hardening were kept the same, the ice creams 
were placed on sale for a period of 10 or more days. 
The ice cream was sold daily to the purchasers, who did not know 
what variation had been made in the composition. Each purchaser 
was provided w^ith a weekly coupon ticket, entitling the recipient to 
one-half pint each day. 
The purchasers were given a sample of each kind of ice cream on 
the first day in order to compare the three kinds, and were informed 
