When the question of statistical significance is asked, one is in actuality asking whether some observed difference between responses of two or more sample segments is large enough such that it is unlikely the result of random chance.  Statistically significant differences then prompt managers to invoke differential treatment, in terms of communications, positioning, and targeting strategies, with the differing segments of the market.

Whether an observed difference is statistically significant will depend on:

• Size of the difference between the groups
• Sample size of each group
• Variability within each group as measured by the standard deviation of the means

Some of the significance testing techniques used by Statistical Reasoning include:

• One-samples t-test of means
• Independent samples t-test of means
• Paired-samples t-test of means
• Z-test for proportions
• Chi-Square test
• Analysis of variance

Caution, however, is always warranted when interpreting differences, be they statistically significant or otherwise.  While it is obvious that a mathematical difference may not be statistically different, a statistically significant difference may not always have managerial or practical significance.  For example, some statistically significant differences are too small to be important.