There are two reasons for our choice of only examining the relations between perceptual
variables. The ®rst results from the way in which we set out our hypotheses on the eciency of
compensation in relation to satisfaction and to motivation. The ®rst three hypotheses, relating
to the concept of motivation, seek to answer the question: `is an employee who is incited by
this or that component of his compensation more motivated to perform at work?' The fourth
hypothesis, relating to the concepts of pay satisfaction and job satisfaction, seeks to answer the
question: `is an employee who is satis®ed by this or that component of his compensation more
satis®ed in his job?' Thus in each case, we examine whether a ®rst ®eld of perceptions has an
in¯uence on the second (see the encircled parts of Figure 1). If pay satisfaction has an in¯uence
on job satisfaction, and if being incited by compensation actually increases motivation to
perform at work, then the compensation, or one or other component of the compensation, is
ecient.