This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Prediction for the election of 1936The Digest vs. Gallup
Roosevelt winsRoosevelt wins
Landon winsLandon wins
Roosevelt winsRoosevelt wins
Gallup’s overwhelming victoryGallup’s overwhelming victory- Serious problems in sample selecting methods of Serious problems in sample selecting methods of the Digest.the Digest.- Gallup even predicted the wrong prediction of the Gallup even predicted the wrong prediction of the DigestDigest
Sample Selecting Methodsof the Digest :-Survey using address from a tele-phone directory or club member lists-In those days, telephone or clubs were the symbols of the rich-the poor were excluded from the sample-The Digest got unfavorable results to Roosevelt who is popular to the poor
-Systematic tendency(Sample selection bias) is related to sample selection. -The sample was not selected in a fair way. -The sample does not represent the population
If there is sample selection bias, the sample doesn’t represent the population. The problem is not solved just by increasing the magnitude of the sample.
If there is sample selection bias, the sample doesn’t represent the population. The problem is not solved just by increasing the magnitude of the sample.
-Only 2.4 million people bothered to reply, out of the 10 million who got the questionnaire-lower-income and upper-income people tend not to respond to questionnairesso the middle class was over-represented among respondents.
Non-respondents can be very different from respondents.
(non-respondent bias).
-Respondents do not represent the sample
Non-respondents may differ from the respondents. If the percentage of the non-respondents is high, you should consider the possibility of non-respondent bias.
Non-respondents may differ from the respondents. If the percentage of the non-respondents is high, you should consider the possibility of non-respondent bias.
- Simple random sampling, Multistage cluster sampling - features The interviewers have no discretion at all as to who they interview. There is a definite procedure for selecting the sample, and it involves the planned use of probability.- With the probability method it is possible to compute the
probability that any particular individuals in the
population will get into the sample.
- For this makes the reproduced research by other people
possible, we can construct the grounds for judging the
significance of the results.
- With the probability method it is possible to compute the
probability that any particular individuals in the
population will get into the sample.
- For this makes the reproduced research by other people
possible, we can construct the grounds for judging the
The Gallup Survey predicts very accurately through a small sample using the probability methods.
To minimize bias, an impartial and objective probability methods should be used to choose the sample. But judgment and choice usually show bias, while probability is impartial.
To minimize bias, an impartial and objective probability methods should be used to choose the sample. But judgment and choice usually show bias, while probability is impartial.
Though you use probability method, you cannot exclude all the bias..-Eligible voters who does not vote-Eligible voters who didn’t make decision-Response bias & Non-response bias etc.To solve this problem Gallup made many survey questions.
The Gallup sample usually includes proportionately too many people with higher education.
Below high school(25%) Above high school (75%)Population
SampleLess weight is put on the responses of higher educated subjects. Lowly educated subjects (Below high school)=weight 1, highly educated subjects (Above high school)=weight 3/4
This weighting technique is an objective arithmetic technique applied to the sample after it is chosen to compensate for various small biases in the sampling procedure. But Quota sampling is a method for choosing the sample. It has large, subjective component.
This weighting technique is an objective arithmetic technique applied to the sample after it is chosen to compensate for various small biases in the sampling procedure. But Quota sampling is a method for choosing the sample. It has large, subjective component.
Other factors that can affect political tendency are included.
Some redundancy is built into the questionnaire, so the answers can be checked for the consistency :
Gallup, USA
Gallup, Korea education, political party
, age, sex, job etc.Race and
Religion
Hometown of
grandfather
Inconsistencies suggest the interviewer may not be doing the job properly. A small percentage of the subjects are reinterviewed by administrative staff, as a further check on the quality of the work.
About 1/3 of residential telephones are unlisted.Rich people and poor people are more likely to have unlisted numbers,
so the telephone book tilts toward the middle class.
5. Telephone Surveys
Gallup : There are 4 time zones in the U.S. The Gallup Poll divided each zone into 3 types of areas, according to population density. That gives 12 strata. Within each stratum, the Gallup Poll just drew a simple random sample of telephone numbers, using random digit dialing.
Telephone Surveys
Telephone Surveys
Non-respondents create problems, as usual.Gallup: Most of its interviewing on evenings, and the weekends,
when people are likely to be at home.
The cost is about 1/3 as much and the predictions are fairly well on target.