SAMPLING CHALLENGES 2
Sampling Challenges
With no doubt, reliable polling requires the sample participants to match the features of
the population where the researcher draws the sample. As it has been the usual case, it is not
practical to interview anyone in an entire community due to time, resources, and availability of
people among other factors. Therefore, there is a justification to pick a few to represent others
hence the need for them to resemble the rest in all aspects. Nonetheless, to achieve this objective
is not easy as it sounds. Reason being, it is very challenging to get a real sample that gives
accurate information that can predict the behavior of a larger group (Zukin, 2015).
Apparently, this is a serious problem that no researcher should ever overlook. One of the
main reasons behind this assertion is the fact that it can lead to sampling error since the sample in
use does not accurately represent the population of concern. Consequently, issues with data
validity and reliability surface to the extent of producing misleading pollsters (Desilver &
Keeter, 2015). A perfect example is where some predict a very close race between two
competing candidates, but it turns out to be very different after polling. More so, with the issue
of unresponsive bias being dominant as well, it is justified to deduce that the problem is severe
since it leads to erroneous surveys and it should never be the case.
In response, several solutions can address the problem. With the main problem being the
type of respondents, the interviewers must always ensure that the sample population contains the
people they are reaching (Desilver & Keeter, 2015). Another possible solution is taking a larger
sample since it improves accuracy and also addresses the problem of low response rates
(DeVault, 2017). Above all, it is vital to carry out every pollster with not only educated people
but also highly trained to understand people’s characteristics.
