Consider an example in which you are studying the business lifespan of shops that have ceased trading in a particular shopping complex and only ten shops have closed down since the shopping complex was opened. You feel that you have the time to study all ten of them. Your study will, therefore, tell you about these ten businesses, and, strictly speaking, no more. You may want to make some qualitative inferences from your study in which you suggest that your study may be representative of failed businesses in general. But, given the way in which you set up your study, namely that you deliberately chose to study a particular shopping complex, any such inferences would have no statistical validity. Whether or not this matters to you will depend partly on the question you are asking, and partly on the approach you have taken to answering it.