Hello. I’m in the process of conducting some preliminary research to assess the impact of a program around various schools in the district. My committee has recommended the use of Hierarchical Linear Modeling (Raudenbush and Bryk (2002) but I am having trouble on recommendations for sample sizes, particularly for Level 1 and Level 2 units. Are there guidelines (or a software implementation like G-Power) to address this kind of situations?
It’s an important research project because a grant will come attached to it so I am trying my best to make sure I have my basics covered and statistics has never been my strongest suit.
This seems to put the cart before the horse (basing a project on a method of presenting data at the end). A more important issue would be the proposed design. How are you going to assess impact? What is the counterfactual? How are cases allocated to comparator groups? How many cases are there in total? Will you have before and after scores? Is there any attrition?