ABSTRACT
This part introduction presents an overview of the key concepts discussed in the subsequent chapters. The part addresses the issues of meaning, power, and level by means of several concrete examples and relates these issues to the issues of form, process, and purpose in scientific inference. It demonstrates that the automatic features of inference with significance tests often lead to conclusions that are patently incorrect by any criterion of good judgment and common sense. The part illustrates the sense in which failure to consider the power of a test can lead to questionable conclusions. It shows that the findings reported in published studies actually have a much larger probability of unsuccessful replication than is implied in the significance level reported. The part suggests that the publication of a significant finding tends greatly to decrease the probability of the study’s replication in the future.
