“There is no doubt that this book will be well received by those who are fortunate enough to come across it. This book will be of use to the growing number of people involved either as purchasers or providers of research. Don't go to work without it!” – Health Services Management Research Journal “I would recommend [this book] to a colleague as a useful companion text for students. I would say that this is an engaging discussion of experimental research for social, behavioral, and health science students. The writing style is fresh and entertaining, and draws the willing reader into thinking through the process of designing and conducting experimental research. It is not a ‘cookbook’ or a compendium of facts. Rather, it is a pragmatic and thoughtful description intended to help students understand how to design meaningful experiments, and by understanding that, they will also understand how to interpret research they do not conduct themselves.” – Katharyn A. May, School of Nursing, Vanderbilt University “This slim but packed volume is written for prospective researchers in the social and health sciences. The writing style is lively, encouraging, upbeat. R. Barker Bausell brings science down to earth without sacrificing respect for rigor and complexity…. Recommended for all institutions with undergraduate or graduate research requirements in the social and health sciences.” – Choice Tired of research methods books that tell how to perform a research study without any mention of the why behind doing research? Aimed at communicating the excitement and responsibility of the research process, this remarkable volume enables you to evaluate beforehand whether a prospective research study has the potential to either improve the human condition, contribute to theory formation, or explain the etiology of a significant phenomenon rather than to produce just another “publishable” study. By emphasizing how to think about and strategize a research study, R. Barker Bausell shows you the important steps of a scientific study–from the formulation of the problem to the write-up of the results. Replete with illustrative examples drawn from the social, health, and behavioral sciences, this volume is a must for all serious researchers.
Designing Experimental Studies to Achieve Statistical Significance
The previous chapter was dedicated to avoiding results due to factors extraneous to the experimental intervention(s) being tested. As we discussed, such erroneous results may be in the form of either:
- false positive results (e.g., the intervention may have been completely ineffective, but more subjects with poor prognoses dropped out of the experimental group, thus making it look beneficial) or
- false negative results (e.g., the intervention may have been quite beneficial, but poor prognosis subjects dropped out of the control group, thus inflating its overall mean and thereby masking the true experimental effect).
There is one genre of error, however, that can be problematic even in studies that follow all the advice presented ...