Discussion
Student ice cream preference
Our study contained several types of biases as well as
a few confounding variables. Not all students who were given
the survey responded, which caused voluntary bias to exist
in our study. This type of bias might’ve caused only those
who had stronger opinions one way or another to answer the
study, thus creating an inaccurate study. One weakness in our study
is
that we had a small sample size, there was a difference in
our proportions but potentially due to the small sample size
we had a large p value and could not reject our null
hypothesis. Another weakness in our
survey is that some people may have no preference and just
circle the first flavor, chocolate, changing our data.
There is significant nonresponse bias in our survey,
only about half the people contacted responded. We
do not feel confident extrapolating the data, we surveyed
only students at NOHS and we cannot assume that the
proportions for something like ice cream preference would
accurately reflect the proportion for other populations.
To gather more information we could send out more
surveys to gather a larger sample, if the proportions stayed
the same with a larger sample our p value would go down and
we might be able to conclude that a statistically
significant difference existed. We could
also survey a different population, such as people in North
Olmsted this would help us understand if there is a
difference between the preferences between males and females
for that population.