THE CASE AGAINST STATISTICAL SIGNIFICANCE TESTING REVISITED