Lesson 6

Lesson 6 Summary

In statistics, level of confidence is used to show how confident statiticians are that the data captures the true meaning of whatever is being tested. It also gives an indication of whether or not the results are repeatable. In other words, if the survey or test was repeated, level of confidence measures how certain you could be that you would get the same results over and over again.

BananaStock/Thinkstock

The level of confidence of the data or the certainty of the results can be determined by calculating the margin of error, confidence interval, and confidence level of the data.

When survey companies publish results, information that indicates the certainty of the results is usually included. The confidence interval is expressed as the survey or poll result (e.g., 60%) plus or minus the margin of error (e.g., 3.1%). The confidence interval is the range of the mean of a population for a given confidence level.

Confidence level is how confident you are that the test results would fall within the confidence interval if you repeated the test. A confidence level of 95% tells you that if the survey was repeated 20 times, the survey result would fall within the given confidence interval in 19 out of those 20 times.

When comparing the level of confidence to the normal curve, the area under the curve represents the level of confidence. The 68-95-99.7 rule relating to normal curves and standard deviations can help you to find a 95% confidence interval quickly on a normal curve. By finding the values of the boundaries for this region (within two standard deviations of the mean), you are identifying the confidence interval. Since 95% of the data will fall within two standard deviations of the mean, you can state that you are 95% confident that the results are between μ + 2σ and μ – 2σ.

In Lesson 7 you will look at what types of factors affect level of confidence and what types of things you need to consider when making inferences from statistical data.