The very limiting factor for this experiment is the sample size. Although ten samples and an overall population of 398 are obtained, the price distribution for local and imported
apples is very restricted. Within the four price levels included in this research, imported
apples contain only two price levels (0.68$ and 0.99$/lb) which make
correlation analysis impossible. Other possible errors may arise from the
deliberately chosen location and time period for data collection. Consumers in the
Lanley market may not be representative of the population in Richmond, and the
time of data collection may exclude certain populations. For example, people
who work in Lansdowne may be unable to shop during the same period when
the experiment is conducted. Some consumers from Delta and Burnaby may also
wish to shop in Lanley’s Farm Market on Saturday. Furthermore, only Friday and
Saturday are included in the data, therefore the data may exclude consumers who
wish to shop on Sunday and weekdays. Future studies may expand the data
collection procedure to include weekdays and record for longer time periods.
However, errors in population representativeness may be hard to eliminate.

Source:

Iggers, D. (2008). Royal Gala apple. Retrieved April 4th., 2012, form http://www.flickr.com/photos/fortinbras/3015463951/