Statistical significance tries to verify that variables related to an outcome are relevant to that outcome. For example, a finance engineer wants to know if a set of stocks will drop within the next 120 days. His model consists of several variables, such as earnings, technical indicators, and news events. Certain news events show a high correlation with stock price movement. The news event variable is, therefore, statistically significant. Any variable that is statistically significant has a high percentage (i.e., close to one). 95% and 99% are commonly used to show statistical significance.
When analyzing a population, most data analysts will use a sample size. From there, they can determine statistical significance. However, it is important that the sample accurately represents the larger population. Otherwise, any statistical significance findings may be incorrect.