The purpose of this investigation was to study the
efficiency of units of practice when unaccompanied by
reward. The method devised was that of running two
groups of rats through the maze: an experimental group
which received no reward during the first part of learning,
but which suddenly had reward introduced in the latter
part of learning, and a control group which received
reward throughout the whole of learning. The answer
to the question as to the efficiency of non-reward units
of practice was sought in a comparison of the learning
curve of the experimental group (both before and after
the introduction of reward) with that of the control
group.

Most of the previous experimental work on rewards
and their relation to learning has been concerned with a
comparison of the effectiveness of different incentives
as such. The incentives have been sometimes different
in quality as food and escape, and sometimes the same

-39-