In this tutorial, you're going to learn about survey applications. You’re going to learn about two companies that use surveys very often in order to obtain useful results:
So let's give a little bit of background into the Gallup organization.The Gallup organization began in 1936 to try and predict the outcome of the 1936 presidential election. It was Franklin Roosevelt, who was the incumbent, against Alf Landon, the Republican from Kansas. Now, there was a magazine out at the time called the Literary Digest. And Literary Digest had a very good reputation, very reliable reputation, for picking the winner of such contests. In 1936, they took on their most ambitious poll yet. They predicted that Landon would win in a pretty big landslide, 57% to Roosevelt's 43%. They did this by having 2.4 million of their subscribers return a questionnaire about who they intended to vote for. This is one of the largest samples and opinion polls that has ever been taken in the country.
Contrast that with what Gallup decided to do.
This is George Gallup. For his survey (or as it's now known, Gallup Poll), he took some of the Literary Digest results, only 50,000 people. But he carefully selected those people from among the 2.4 million respondents to make sure that his group of people was demographically representative of the country as a whole, as opposed to just counting everyone who chose to send their ballot back (as the Literary Digest had). His prediction was that Roosevelt would win 57% to 43% over Landon, essentially a reversal of Literary Digests numbers.
Polling method developed by George Gallup whereby a very representative sample can be taken with a fairly small sample size.
So who was going to be vindicated? Well, it ended up being one of the most lopsided victories in American history.
The blue counties here are counties that voted for Roosevelt. And the red counties voted for Landon. You can see that almost a large majority of the counties voted for the Democrat, and Gallup was vindicated. His process of selecting a representative sample as opposed to those who volunteered to return their ballots ended up being the gold standard for sampling.
Gallup's organization is still working today. They do polls for all sorts of things: opinions on presidential approval ratings, who will win in an upcoming presidential election, etc. And they still use a sampling technique and surveys to try to obtain representative samples for the entire country. Have you ever answered or come across a Gallup poll in your life?
The Nielsen company deals with entertainment trends in America, by surveying public interest about different media and retail products or events. They release this information in what we call the Nielsen Ratings.
A public interest study that measures mainly retail and media consumption.
To perform these ratings, the Nielsen Company takes a small sample, and small meaning several hundred or a few thousand households in the US. But by taking a representative sample, they can determine the television viewing and radio listening habits of Americans with a good degree of certainty.
This research happens in two ways:
Using these two methods can pretty accurately determine who is watching what, or who watched what in any given time frame.
So if these ratings are so useful for TV and movies, what about music on the radio? In order to determine public interest in radio programming, the Nielsen Company also performs the Nielsen Radio Index, which uses a very similar method to the Ratings.
Nielsen Radio Index
A method for measuring radio listenership that was first used in 1942.
Surveying is done extensively. The most prominent examples of surveys and large scale sampling are done by the Gallup polling organization and the Nielsen market research company. Gallup deals with politics and social trends, while Nielsen deals mainly with entertainment, both in TV and on the radio.
Thank you and good luck!
Source: THIS WORK IS ADAPTED FROM SOPHIA AUTHOR JONATHAN OSTERS