Question

• What do you think it means to say that a test is “statistically significant”? How can saying this be used to strengthen inferences drawn from such a test?

• Can you think of an event that occurred in your personal or professional life that required statistical significance? Share it with your fellow students.

Answer 1

We say that a test (rather a decision) is 'statistically significant' if the null is rejected in favor of that decision as an alternative. What it actually means is that we have enough evidence from the sample to support the decision. As an example consider, giving drugs to patients that claim to speed up recovery rates. Now, some patients might 'inherently' have a higher recovery rate. These things are called random errors. To say that the effect of the drug is 'statistically' significant is to say that the drug really has a higher recovery rate of it's own, and the higher observed rates are not due to random errors. Thus, this gives us a confidence, that the decision is 'really' true (with low error rates) and hence, the it strengthens the inference.

An example maybe this: Assume that you started doing some computer intensive jobs and your provider guarantees that the computer doesn't consume much electricity. So in order to verify that, you can perform a test to see whether the increase in electricity consumption before and after the advent f computer is 'statistically significant' or not.