Sources of Error
There are many sources of error in any data collection activity. Some of the error or bias can be overcome. For example, some sampling error may be overcome by weighting the data to reflect the appropriate size of certain subgroups in the population. Weighting is used when the random sample comes up short in a particular group, say young men, 18-24. The results of the young men in the sample might be weighted-counted more than once per respondent-in order to secure the correct proportion to be representative of that group nationwide.
For a discussion of sampling error and margin of error, as well as these other types of error or bias are described in Polling Fundamentals section:
Errors in designing questions are at least as frequent as any of these. Always obtain the question wordings and report them when analyzing data. See the section on Questionnaire Design.
For further information please contact The Roper Center at 607.255.8129 or email@example.com.