Significance Testing
In the medical or nursing research, it is more common to use the term significance testing whereas in the statistical research, the term hypothesis testing is used more often.
Significance testing refers to the use of statistical methods to make categorical conclusions about the population of interest, using the data at hand (sample). Specifically, we formulate the null and the alternative hypotheses about population and determine whether or not we have sufficient evidence to conclude the alternative hypothesis can be established.
Example 1
Suppose a stress indicator, ranging from 0 to 100, is used to examine the stress level in Hong Kong nursing students. To examine if the mean stress level is over 50, we may perform a one-sample t-test on a sample of subjects, provided the stress score follows a Normal distribution. The conclusion can be either sufficient or insufficient evidence that the population mean stress level is over 50.
Example 2
Suppose we are interested to examine if the preference of capturing laughers with a camera is associated with gender. We may randomly select a sample of subjects and perform a chi-square test. The conclusion can be either sufficient or insufficient evidence of an association.