You’ve invested an extensive amount of time and money to create your survey and collect responses. But none of it means anything unless you’re able to glean meaningful results. One way to do this is by calculating statistical significance.
Statistical significance refers to a result that is not likely to occur randomly, but can instead most likely be traced back to a specific cause.
Random Results or Specific Cause?
Perhaps you find survey respondents who live in Alabama are absolutely crazy about your products, but those who live next door in Georgia rarely buy a thing. Statistical significance can ascertain if there’s a likely cause behind the dramatic differences in purchases between the two adjacent states, or if it’s a totally random event.
Your statistical significance can be weak or strong, and you determine its strength through significance testing. Significance testing is a key tool for determining the reliability of survey results.
Significance testing is always subject to a certain margin of error. Unless your survey includes the entire population, which is not possible for any survey, you have to account for a sampling error. The larger your sample, the less prone your results are to flukes.
What Significance Testing Entails
Researchers typically use statistical hypothesis testing to determine if a specific result is statistically significant – or not. The testing provides a “p-value,” which is the term used to describe the likelihood that the outcome is the result of random chance. Generally, a p-value of 5 percent of less is considered statistically significant.
If you found your Georgia-Alabama purchases provided a p-value of 50 percent, for instance, it would be likely that the outcome was due to chance and not statistically significant. If the p-value was 2 percent, on the other hand, there is likely a cause behind the differences and the results would be considered statistically significant.
Where Crosstab Software Comes In
Market researchers were once forced to use complicated, costly software that required extensive expertise and training to determine statistical significance that pinpointed the most important survey findings.
But you can get the same level of accuracy in a fraction of the time with crosstab software. Instead of requiring costly software and extensive expertise, quality crosstab software automatically performs the statistical testing for you.
- Automatically configure and run applicable statistical significance tests
- Allow you to configure different options if desired
- Alert you to statistically significant findings
Now let’s say your crosstab software did give you a p-value of 2 percent for our Alabama-Georgia results. You would now know there was most likely a traceable cause behind the dramatic differences in purchases, a fact that could lead you further investigation and subsequent strategies to boost your business across Georgia – as well as the entire nation.