The most effective techniques at gathering information from an audience are polls and surveys. In the United States there are about 300 companies which specify in surveying people for private clients, these include Nielsen Media Research, which specializes in network television ratings, the Pew Research Center which studies attitudes towards the press and politics, and Arbitron, which mostly measures radio audiences in local markets. By using the results from these surveys, a good public relations officer can figure out the best time to place an ad for a specific product. In the area of polls and surveys, there are even smaller branches which are specific ways to poll and survey. One of these methods, probability sampling, relies on sample size, sample selection, and confidence level to determine a certain group's reaction to a question. The magic number for sample size has been determined to be 384, meaning that no matter how large the size of the group you are polling, 384 is the maximum amount of people you will need to directly survey. The next step is determining the sample selection, you do this by taking the total population and dividing by the magic number, and that number is the interval between people you should survey. Lastly, confidence level is inversely related to the margin of error, as a larger sample size can increase confidence and decrease margin of error.
The other method of surveying is quota sampling, where you figure out a ratio between, for example, the number of men in the population size, and the size of your survey group. You then interview that many men to fill your quota and repeat for any other group in your population. This method was used by Gallup to come up within 4 points of the average but then switched to probability sampling to come up within 2 points of the average.
No comments:
Post a Comment