Unit 3 Individual Project

Introduction

Polling presents policy makers and other interest groups important insights about public
opinion on various issues. For policy makers, results of an opinion poll may encourage them to
implement a given policy, modify it or drop it altogether if an overwhelming majority of the
public is opposed to it. These reactions to opinion poll findings are based on the assumption that
they represent an accurate reflection of public opinion on a given issue. This article looks at two
opinion polls by Pew Research Centre, analyzes their accuracy based on sample size, and
examines how their findings could be used by various interested parties to pursue their goals.

Political survey

I chose the poll titled “Public Expects Gridlock, Deeper Divisions With Changed
Political Landscape”. The poll sought to find public’s opinion on how the results of the midterm
elections of November 2018 were going to affect the ability of Congress and the president to get
things done on a wide range of issues. The poll also sought public opinion on who between the
president and the democratic-controlled House the public believed should lead America on issues
such as the environment, job creation, foreign policy, gun control, trade, and immigration (Pew
Research Centre, 2018).

I chose the poll because the mid-term elections in the US produced mixed results. On the
one hand, Democrats were handed control of the House of Representatives. On the hand,
Republicans were allowed to retain control of the Senate. Thus, with such mixed results I
thought it important to use a political poll that clarifies what the public really wanted and their

Running head: Project Type: Unit 3 Individual Project 3
expectations from their new Congress as well as the presidency. I was also attracted by the poll’s
relatively large sample. At 9,451, the sample size of the poll was among the largest among all
polls.

The poll, conducted across the nation, uses a sample of 9,451 adults. American adult
population is approximately 230 million. The sample, therefore, represented 0.004% of the adults
in America. This number might look like a very small portion of the population and, therefore,
unlikely to yield relatively accurate findings. However, the total adult population in America has
little bearing on the accuracy of the findings obtained from the sample (Shively, 2016). Its
accuracy (or lack of) would remain the same whether the overall adult population was 230
thousand, 230 million or 2.3 billion. What matters is the confidence level and margin of error
(Peck & Devore, 2015). A sample of 9,451 has very high confidence level. It also has low
margin of error. Thus, to that extend it is a fairly accurate representation of public opinion.
However, it should be noted that America is a very diverse country with very sharp political
divisions. Thus, it would be useful to know the extend to which the sample represented the
various political groups. For instance, it would help to know the percentage of participants that
self-identified as Democrats or Republicans. Knowing the political affiliation of the participants
would help determine whether there is bias in the sample or not. In other words, the sample size
itself is not the problem; the problem is how representative it is.

The results of the poll show that a majority of Americans expect gridlock in Washington
following the election of a Democratic majority in the House. Equally important is the finding
that shows that Americans trust the Democratic-led House to provide leadership on all but two
issues surveyed. The issues include healthcare, environment, foreign policy, and government
ethics (Pew Research Centre, 2018). President Trump only leads on the issue of jobs. For

Running head: Project Type: Unit 3 Individual Project 4
Democrats and their backers the poll findings may be used to legitimize their policy initiatives.
They will argue that Americans prefer their policies over those of President Trump. They will,
therefore, have more confidence when pushing for their policy initiatives because they know that
they have the support of majority of Americans.

The President and his Republican backers will find solace in the fact that the poll shows
Americans have very high confidence in his ability to create jobs. However, given the findings
are overwhelmingly not in their favor, they may argue that given that Democrats have yet to
come up with very clear initiatives, it does not make sense to find out public opinion on the
usefulness of their policies. Republicans may, therefore, agree with the findings of the polls but
may dismiss them as being premature.

Environmental Survey

For environmental survey I chose the poll titled “Most Americans say climate change
affects their local community, including two-thirds living near coast”. The poll was published on
16th May, 2018. The survey sought to find public opinion on the effect of climate change either
in their personal lives or on their communities (Kennedy, 2018). The sample was fairly
representative with both Republicans and Democrats involved as well as people living along
coastlines and inland. The sample size of the poll was 2,541 (Kennedy, 2018).

I chose the poll because it is relatively recent. Its findings are, therefore, relevant in the
current discourse about the environment and climate change. Additionally, it tackles an issue that
has sharply divided public opinion in the U.S. One of the hurdles that has prevented America
from taking lead on matters regarding climate change has been lack of consensus on whether

Running head: Project Type: Unit 3 Individual Project 5
climate change is real or not. The poll’s findings are, therefore, important in gauging direction of
public opinion on the subject.

The survey used a sample size of 2,541 adults out of the total US population of 230
million adults. The sample might appear small but I believe that it is large enough to produce
fairly accurate results. Most surveys on various issues in the US use a sample size of between
1,000 and 3,000 adults (Shively, 2016). The 2,541 number is, therefore, within the normal
sample range. In any case, as previously explained, the is no direct relationship between
population size and sample size. The 2,541 number is just appropriate for 230 million people
just as it is for 100 thousand people. Given the question of the survey, a figure of 2,541 is going
to have very high confidence level and relatively small margin of error.

Thus, I believe that the findings present a fairly accurate reflection of the views of
Americans regarding climate change and its effects on communities. I use ‘fairly’ because the
sample size used cannot produce 100% confidence level. At most, its confidence level is 95%.
Thus, the likelihood of the results not reflecting the views of Americans is just around 5%.

The results of the survey show that two thirds of Americans living near coastline believe
that climate change has affected them personally or affected their communities (Kennedy, 2018).
These results will certainly embolden environmental activists who have been advocating for
stronger measures from the government to combat climate change.

Those opposed to the idea that climate change is real or that government should not take
measures to combat it will argue that the findings do not mean much. Given that most believers
of climate change and its effects on populations are Democrats, and that Democrats tend to
mostly live in major cities along coastlines, they will argue that the results of the polls do not

Running head: Project Type: Unit 3 Individual Project 6
present objective views of Americans regarding climate change. Rather than showing that people
living along coastlines experience the effects of climate change, they may argue that the
concentration of Democrats along coasts makes the poll inherently biased.

Overview

The two polls used vastly different sample sizes for the same population. The sample size
of the first poll is around 7,000 bigger than that of the second one. This, however, does not mean
that the first poll is more accurate than the second poll. It, however, has a lower margin of error
and a much higher confidence level. For accuracy, we may need to also analyze how
representative the samples were given that policy differences also reflect political and
geographical differences of respondents.

Conclusion

As this analysis shows, the sample size of a poll has a huge effect on its margin of error
and confidence level. However, it does not have much bearing on a poll’s accuracy.

Running head: Project Type: Unit 3 Individual Project 7

References

Kennedy, B. (2018). “Most Americans say climate change affects their local community, including two-
thirds living near coast” Pew Research Centre. Retrieved from http://www.pewresearch.org/fact-
tank/2018/05/16/most-americans-say-climate-change-affects-their-local-community-including-
two-thirds-living-near-coast/
Peck, R., Olsen, C., & Devore, J. L. (2015). Introduction to statistics and data analysis. Cengage
Learning.
Pew Research Centre (2018). “Public Expects Gridlock, Deeper Divisions With Changed Political
Landscape pew research centre”. Pew Research Centre. Retrieved from http://www.people-
press.org/2018/11/15/public-expects-gridlock-deeper-divisions-with-changed-political-
landscape/
Shively, W. P. (2016). The craft of political research. Routledge.