Blackwell Reference Online

Extract

A confidence interval (CI) is an interval estimate of a population parameter. It is a range of values, calculated from data, that is likely to include the true value of the population parameter. When a newspaper reports “support for the government is 43 percent, in a poll with an error margin of 3 percent,” the 43 percent is a point estimate of the true level of support in the whole population. The CI is 43 ± 3 percent, or (40, 46 percent). The 3 percent is half the width of the CI, and is called the margin of error . The endpoints of the CI are the lower and upper limits or bounds . The level of confidence , C , is expressed as a percentage. Most commonly, C = 95 is chosen, to give 95 percent CIs, although 99 percent CIs, 90 percent CIs, or CIs with other levels of confidence may be used. Understanding level of confidence is the key to understanding CIs, and will be discussed in the context of an example that also illustrates calculation of a CI in a simple case. To estimate μ, the mean level of community-mindedness, we administer a measure to a random sample of n = 30 people from the population, and calculate mean M = 59.52 and standard deviation s = 32.4. The margin of error is w = t C x s /√ n = 12.11, where t C = 2.045 is the critical value of t , with ( n – 1) = 29 degrees of freedom, for confidence level C = 95. The 95 percent CI for μ ... log in or subscribe to read full text