What does statistically significant mean?
Statistically significant means that the difference between two numbers is large enough that it is very likely, 95 percent of the time, that the difference would appear again if a completely different sample of respondents were asked the same question. At the 95 percent confidence level, there is only a five percent chance that the differences observed do not reflect actual differences in the total population of interest, but rather are due to chance or sampling variation. We test for statistical significance generally using a Z-score or a T-score. The Z-score is used to test the difference between two proportions or percentages, while the T-score tests the difference between two means. You can conduct these tests yourself using the Basic Statistical Testing page, accessible from the Reference Center in the left column on the StatTrac home page or from the Visitors Center at www.polarismr.com.
