Different Ways to Reduce Margin of Error
The media has coined a new jargon called margin of error and is commonly applied in survey results analysis. The margin of error is used to quantify or attach a numerical value to the uncertainties associated with a survey result. In other words, sampling errors are commonly summarized by this term. Margin of error is similar to the laws of chance or the laws of probability. The objective of survey takers and poll conductors is to reduce the margin of error to get a more accurate result.
Other People Are Reading
-
Sample Size
-
The three factors affecting the margin of error are the size of the population, the type of sampling done and the sample size. In a telephone poll, if 100 people are polled, the margin of error is almost 10 percent. If the sample size (number of people polled) is 500, the margin of error decreases to 4.5 percent approximately and a 3 percent margin of error is produced when the sample size is 1,000. Thus increasing the sample size minimizes the margin of error. If a 1.5 percent margin of error is required, more than 4,000 people have to be polled.
Size of Population
-
This factor has no telling effect on the margin of error. If the size of the population is 10,000 and the sample size is 100, the resultant margin of error will be the same if the sample size was 100 and the size of the population was 10 million. Thus the size of the population does not influence the margin of error much.
-
Type of Sampling
-
Stratified sampling, random digit dialing and simple random sampling are the three common types here. When a survey is being conducted to get an estimate of the differences in viewpoints among dissimilar groups, using stratified samples reduces the margin of error. To reduce the margin of error (in case of a census), the interviewer has to visit many blocks or sample many cities or many counties. Decreasing the sample size increases the margin of error in this case as well. Simple random sampling is done with the help of computers to reduce the margin of error.
Reducing Margin of Error in a Poll
-
When greater numbers of people are polled, the margin of error is minimized. Professional polling agencies are roped in to design political polls and they involve only sufficient numbers of voters so that the margin of error is tiny (it is two to three percent usually). To restrict the margin of error to three percent, a little more than 1,000 people are typically required. If fewer people are polled, the results would be less accurate with a higher margin of error. Polling more people would not be feasible, as it is expensive.
-
