No doubt that presents a very informative paper on why and how to define sample sizes. I recommend reading section 7 ‘Avoid Retrospective Planning’. An interesting insight involves his sentence concerning the idea that to obtain statistical significance, while ignoring scientific meaning, it is just asterisk hunting! Important for those supervisors and examiners of PhD dissertations that recommend to increase data collection in order to achieve scientific breakthroughs.

Follow this link: Lenth (2001) text