Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks

The Evolution of Tracking Public Opinion

A presentation on the shift of polling techniques from a mass to a networked society

Lindsay Augustyn

on 15 April 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of The Evolution of Tracking Public Opinion

Tracking democracies accountable modern scientific poll Polls still in the experimental stage' Gallup survey finds public is less certain in 1985 than in 1944 about the reliability of polls on issues relative to election predictions - Andrew Kohut Critical turning point write a walk How did Literary Digest get the prediction of the 1936 presidential election so wrong? At first, statisticians blamed the uneven distribution of telephones as the culprit for Literary Digest's prediction that Kansas governor Alf Langdon would beat Franklin Roosevelt. The magazine correctly predicted the winners of the 1920, 1924, 1928 and 1932 U.S. presidential elections. However, it met its demise with its incorrect prediction of the 1936 presidential election. 'While public opinion polls are today sometimes viewed as part of the problem of political life, ... in 1940 they were very much a part of the solution.' Public Opinion The Evolution of Tracking Role of public became key to keeping Straw polls first began with the magazine Literary Digest. Developments in scientific sampling lead to However, a study by Dominic Lusinchi published in 2012 in Social Science History declared that since only 2.3 million postcard ballots were returned of the 10 million that were mailed, non-response bias was to blame - in other words, Landon supporters were far, far more likely to fill out the postcard and send it in.

George Gallup's American Institute of Public Opinion, on the other hand, correctly predicted Roosevelt's win and gained national recognition. After the accuracy of the Gallup poll, there was no turning back. More than just political campaigns; Clicks and trending online matter Rise of era of modern scientific public opinion research 20th century + From Postcards to Clicks:
Shifting from Mass to Networked Communication COMM 852
Lindsay Augustyn
Final Project
Spring 2013 21st century 'The A lack of confidence in polls followed in 1984 "They [Gallup and Rae] obviously represent a new era of study, which is defined by a more scientific, quantitative notion of public opinion; yet, they also painted a broad normative and theoretical strokes in a way that harked back to the earlier students of opinion such as [Walter] Lippmann [in 1922's Public Opinion in which he stated that the public is not competent enough to be polled] or Lowell. The opinion poll was an aid in a less than perfect democracy. It was not a substitute for the citizenry." Gallup represents earlier and later periods of opinion research: Michael J. Korzi, "Lapsed Memory? The Roots of American Public Opinion Research" The final pre-election survey in 1948 was in mid-October when it was felt that opinions had solidified. Obviously, a large error in prediction ensued [Dewey defeats Truman]. This led to Gallup making changes to their survey procedures in 1950. - Paul Perry For more than two decades, most of the national polls had produced final figures that were either right on target or well within sampling error. But, in 1980, the polls failed to detect President Reagan's big victory. - Andrew Kohut, president of Gallup, 1986 Leaders pay attention to group opinions, especially powerful ones. Because sampling methods used by pollsters do not differentiate along group lines, they are not up to the task of observing public opinion. - Herbert Blumer Illuminating the deficiencies of
standard public opinion research Foreground Middleground Background Advent of that record the candidates' national standing on a DAILY basis 'Social 'polling' exploring new ways to track What is the best way to track public opinion today? Online poll aggregators Still... when polling and social media collide public opinion today 2012 Presidential Election Weighting adjustments Semantic analysis Tracking polls provided content for news stories and campaign tacticians about how potential voters are responding to various campaign events Horse race journalism and news polls, courtesy of The Daily Show Local news "Polling fatigue" blah blah blah blah blah blah blah blah blah blah blah blah There has consistently been a bias toward describing polls that appear in newspapers as opposed to on television because of greater ease in assembling information from content analyses of such sources. Phones Tracking polls How estimates made Criticisms Like a floor plan...
It makes sense from the overview Horse race journalism Sampling error A threat is growing to the accuracy of election polling in general Polling organizations started experimenting with weighting adjustments for party identification if your prezi looks like this, please work a little more, scale up main points, place subtopics into larger frames, etc... Could the
exclusion of
cell phones
really make that
much difference? Twitter & Twindex exist with social media and online polling Online polls provide biases because not everyone has access to the Internet Differences in tracking poll results suggest more scrutiny needed on methodologies Here are some of the issues we face: Social polling may not have the same accuracy or reliability as random sampling, but it has tremendous promise.
Independent pollster Zogby believes 92% of likely voters will be online someday soon. flaws and biases Some ways of using social media to track public opinion in 2012 election The rapid growth of cell phone use raises concerns
about undercoverage error in telephone samples. Now, roughly one-third of U.S. households
do NOT have a landline.

Potential voters who rely on cell phones belong to more Democratic-leaning demographic groups than those which don’t, which skews polls to the favor of Republicans. Katie Couric on Cell Phone Polls: How else do we track opinions on social media? Digital divide Semantic pollsters need to look for ways to work on subtleties, coping with things like sarcasm and irony
Most importantly, social network users are not representative of the entire population, according to Zogby. Pew reports that 80% of them are between ages 18-24, while only 26% are over 50
Another new problem: Distinguishing between political and non-political accounts on Twitter. It is important to distinguish between these two types when pulsing public opinion because we want to separate opinion-makers from opinion-holders (the general public). Political Twitter accounts tweet almost exclusively about politics and tweet to promote their opinions. public opinion is study of President Obama got stronger results in polls that used live interviewers AND cell phones in their samples — enough to suggest that he had a clear advantage in the race, which turned out to be a correct prediction. Next: Social Media Problems with Landlines. vs. Cell Phones

Low telephone survey response rates require pollsters to weight data to "rehabilitate" responses by unreachable parts of the voting public.
Phone number lists for cell phones are more expensive than those for landlines.
Cell phone numbers change frequently and using a computer to auto-dial them is illegal.
A person must call cell numbers manually, which is very expensive.

Simon Jackman, Stanford University Likes on Facebook
USA Today's Twitter Election Meter
YouTube video sharing
Commenting online In "The Pulse of Democracy," George Gallup and Saul Rae argue: "Public opinion must be measured. What the mass of people thinks puts governments in and out of office, starts and stops wars, sets the tone of morality, makes and breaks heroes."

There was a need to discover the opinions of the people aside from merely at election-time: "elections can never be the sole channel for the expression of public opinion." Despite 1948
election errors,
other successes
of polls made
their use soar
with politicians
and academics. The lead article of the very first volume of Public Opinion Quarterly was "Toward a Science of Public Opinion" by Floyd Allport, a noted psychologist (pictured). This showcases the shift to the domination of psychologists in the field of public opinion. The 2000 election campaign saw more polls and more reporting of polls than ever before, as well as a wider variety of data collection methods. Do polls need to pick the correct winner in order to be judged accurate or is just being statistically close to the final outcome good enough? ? The pre-election polls in 2000 campaign, were statistically speaking, as accurate as the average polls since 1952.
Given the closeness of the actual vote, they did
remarkably well at estimating each candidate's
share of the vote. Typical case: Tracking poll estimates are
drawn from small daily samples that are
combined every three days.

Between 150 and 350 interviews were conducted daily by different polling operations. Sample sizes were anywhere from 500 to 1,000 respondents. Gallup
Zogby A dangerous exercise... Respondents' reports of party identification may mean different things early and late in the campaign sponsor their own polls,
leading to public perception of
more polls Average response rate in the 2000 exit polls was 51%, compared to 55% in 1996 and 60% in 1992. In 2005, nearly all cell numbers were excluded from random-digit dial samples, thereby excluding cell-only households. In 2004, 20% of young adults (18-24) were in cell-only households, more Hispanic adults were in cell-only households than other race and ethnic groups, and adults who were not married are more likely to live in a cell-only household than married adults. 2010 http://fivethirtyeight.blogs.nytimes.com/2012/09/19/obamas-lead-looks-stronger-in-polls-that-include-cellphones/ Silver found polls by Scott Rasmussen, founder and president of Rasmussen Reports, were off by an average of 5.8 points, a higher number than most. Rasmussen says his polling firm is already compensating for new technology. "We need to find ways to reach people in social media." http://www.cbsnews.com/8301-3445_162-57544827/taking-the-pulse-of-nate-silvers-numbers/?tag=showDoorLeadStoriesAreaMain;SunMoLeadHero Facebook entered into an arrangement with the news website Politico to implement a simple positive vs. negative machine reading of status updates. Researchers are able to characterize debate performances moment-by-moment, analyzing the sentiment of tweets collected during live debates. Social polling companies would say Facebook, Twitter or a blog. GoPollGo Founded by Ben Schaechter, 23:
Asks people things like: Choose which product you can't live without the most: Google or Apple.
ESPN hosts a GoPollGo poll to decide what the theme for Wednesday's SportsNation poll should be. Quipol Founded by Max Yoder, 23
Asks people things like: Would you ditch cable and just use Netflix and Hulu?
Users can also create polls, which have thumbs-up or thumbs-down buttons below the questions. Websites such as Nate Silver's FiveThirtyEight.com, Pollster.com, and RealClearPolitics.com collect, analyze and graph an enormous amount of state and national polling data and help voters to interpret and assess campaign dynamics by providing singular, statistical measures that capture all of the available information daily. Will texting someday enter the mix? It's an uncharted territory for pollsters. Response bias Some pollsters are giving incentives to people for answering polls - does that mean people are giving true data or the answers they think the surveyors want to hear? Twitter ... and the future
of semantic polling Just as people eventually accepted the major shift in the polling world from knocking on doors to calling landlines (and later, cell phones), the U.S. will adjust to using mostly social media and online polls to gather public opinion. Peters & Simonson
Full transcript