Redressing the power and effect of significance. A new approach to an old problem: teaching statistics to nursing students

Many barriers to learning are present when teaching research methods. Devel oping, within students of nursing, the skills of reading and interpreting r esearch reports is vital if the profession is to contribute to the general aim of achieving a sound basis for all health care interventions. This pape r overviews the current move toward evidence based practice, the challenges that are present when teaching research to nursing students and offers an approach to teaching quantitative research that will help students of nursi ng to understand the key concepts that form the basis of inferential statis tics. In this work we argue that the traditional emphasis on probability an d statistical significance needs to be redressed and that effect size and p ower should form the basis of teaching students the concepts involved in in ferential statistics. We argue that introducing students to the key concept s in statistical decision making a particular order, effect size then power and lastly statistical significance, will lead to a better understanding o f Type I and Type II errors. After all, the purpose of hypothesis testing i s to detect a treatment or intervention effect. Power is dependent upon the size of the treatment effect, thus it must be introduced after effect size . Students, we argue, must be able to understand the concept of effect size . We consider this to be a foundational concept that will help to develop a firmer grasp of the decision making processes involved in hypothesis testi ng. Such an approach will form a more logical approach to teaching this sub ject and will allow for the use of real world examples to form the basis of learning. (C) 2000 Harcourt Publishers Ltd.