Although it had previously accurately predicted the winners of elections, The Literary Digest failed to do so in the election of 1936. The main reason was the method of polling they used. The polled their readers, who were wealthy enough to still be able to afford to buy the magazine during the depression. They also polled automobile and telephone user, who at that time would have been affluent. As a result, they did not tap into the pulse of America.
Respondents intentionally lied about their preferred candidate
Back then, polling and political science research were still very new, and not as reliable as they are today. Literary Digest did a mail-in poll, based largely on people who subscribed to their magazine. As it turned out, many of their subscribers were Republicans, and they did not want President Roosevelt (a Democrat) to win. But as anyone who has done survey research knows, a sample must be RANDOM in order to be reliable. Thus, selecting only your readers is great if you want research on opinions about your magazine; but if you want opinions about the election, you need to reach out to a much wider range of people, and have a variety of ages, races, various geographical locations, various political viewpoints, etc. Literary Digest did not do this; that is one big reason why their results were wrong.
Another reason is that asking people to mail in the reply (called "self-reporting") gets you only those people who are called "actives"-- they are passionate enough about the subject to send you a letter, fill out a survey, answer questions, etc. To get a good sample, you need "passives"-- they make up the majority of the public, and they rarely send a letter or fill out a survey, even though they do have opinions. You need reach out to them, and Literary Digest did not do this. They only got results from the most passionately political people, who make up about 7-10% of the average audience. This was another reason their results were so wrong.
The Literary Digest only polled its subscribers, who were wealthier than most Americans, and therefore tended to have a more negative view of some of Roosevelt's policies. FDR won the election with 60. 8 percent of the vote.
The Literary Digest
The Literary Digest was created in 1890.
The Literary Digest ended in 1938.
George Gallup correctly predicted that FDR would handily defeat Alf Landon by using a survey with a relatively small random sample of about 1000 people; meanwhile the Funk & Wagnalls Literary Digest used a sample of more than a million of its readers, who were disproportionately wealthy and supportive of the Republicans, to predict that Landon would win in a landslide, and at first the Literary Digest poll was more widely known.
The literary digest only polled its readers and its readers were not a random sample. Only people who liked to read and had money to spend on books took the literary digest. Thus their poll was biased and did not give an accurate result.
The Literary Digest
The convenience sample
Sampling error. The LD had been drawing its polling sample from 2 groups: automobile registrations and telephone subscriptions. This method allowed them to predict 5 presidential elections - 1916 to 1932 - correctly. By 1936, however, FDR had disenchanted many of the middle class voters - those with cars and/or phones - who had helped him to his landslide in '32. Car and phone owners that year voted overwhelmingly for Landon, as the LD predicted. But the majority of voters were working class people with neither a car nor a phone. It was they the LD ignored - and they who reelected FDR.
Gallup Poll- Reputation for honestyStraw poll- informal poll Literary digest- committed sampling error
Straw polls . "straw poll or straw vote is a vote with nonbinding results. Straw polls provide dialogue among movements within large groups. [1][2] In meetings subject to rules of order, impromptu straw polls often are taken to see if there is enough support for an idea to devote more meeting time to it, and (when not a secret ballot) for the attendees to see who is on which side of a question. " From wiki
The LD had been drawing its polling sample from 2 groups: automobile registrations and telephone subscriptions. This method allowed them to predict 5 presidential elections - 1916 to 1932 - correctly. By 1936, however, FDR had disenchanted many of the middleclass voters - those with cars and/or phones - who had helped him to his landslide in '32. Car and phone owners that year voted overwhelmingly for Landon, as the LD predicted. But the majority of voters were workingclass people with neither a car nor a phone. It was they the LD ignored - and they who reelected FDR.
Because Pepsin is the active form of a protein manufactured in the stomach.