< Previous | Contents | Next >
An idea that is put to the test. In econometrics, experts start with a null hypothesis (i.e., a particular variable equals a particular number), then crunch the data to verify it in accordance with the laws of statistical significance. The chosen null hypothesis is often just opposite to what the experimenter believes.
Statistical significance means that the probability of getting the result by
chance is low. It is most commonly used measure is that there must be a 95 per cent chance that the result is right and only 1-in-20 chance of the result occurring randomly.