http://www.greenbook.org/marketing-research.cfm/statistical-significance-03377
The above article is titled, "What Every Researcher Should Know About Statistical Significance," and it provides good insight to some key questions.
What is statistical significance?
The article boils this down to two important concepts, sampling error and confidence level. These are two key features when deciding whether survey results can be extended to the larger population.
How large a sample is enough?
The article outlines the process of deciding on a good sample size. It starts with deciding what percentage of error is acceptable and then calculating backwards to find the appropriate range of participants. "The sample precision analysis can help a researcher make an informed decision regarding sample reliability." - Data Star Inc.
What statistical tests do I need to know about?
This portion of the article describes both the Z-test of proportions, and the T-test of independant means. Sample precision analyses typically uses the T-test to estimate an appropriate sampling error. One-tailed and two-tailed tests are also mentioned for testing directional difference and the difference of groups.
Leave a comment