The research suffers from the quality of the composition of the sample. By virtue of consisting of
student shoppers alone, the sample was not sufficiently heterogeneous. The limited heterogeneity
in respondents' demographic characteristics could have affected both the nature and the extent of
the predictor variables attaching themselves to the reference price and the reservation price
especially since there were no financial consequences for the respondents participating in the
investigation, which would not be the case in a real world shopping situation. Also it is possible
that the student respondents, while doing regular grocery shopping, were not the "principal
household grocery shopper." If shopping for themselves these student respondents could be
severely constrained financially and currently buy only economy brands. Any attempt to upgrade
20
this type of shopper to a more premium brand will succeed only if the price drop brings the price of
the higher strata brand to the same level as their regular purchase. This investigation however did
not separate respondents who shopped for themselves and for their households and cannot ascribe
conclusively that because of the high salience of this type of respondent in the sample the reference
price in combination with other variables is about equal to the estimated price. This type of student
shopper would also, in their concern for the lowest price they can afford, be likely to have a
reference price and reservation price that are close.