4.3.3 Confidence intervals

Confidence intervals (CIs) provide a measure of the uncertainty of our estimates and of the confidence that we have regarding the true value of the parameter of interest. Confidence intervals are often interpreted as the range of values within which we expect the population parameter to lie within a certain probability (typically 95%). Technically, this is incorrect; the true interpretation of a 95% CI is that if we were to repeat the study an infinite number of times under the same conditions, 95% of the values obtained from these studies would be contained within this CI.

For example, zoom in on the first row of data in the separate PDF file [Tip: hold Ctrl and click a link to open it in a new tab. (Hide tip)] : the best estimate of the percentage of community isolates resistant to ampicillin in 2009 is 39.6%, but the 95% CI is 36.3%–43.1%. This means that if this study was to be repeated a very large number of times, we would expect the percentage of isolates resistant to ampicillin to fall in this range (36.3%–43.1%) 95% of the time.