Evidence-based "scientific" teaching represents a set of new applications of a well-known teacher*researcher paradigm in which faculty apply the scientific method to pose and test hypotheses based on cognitive learning theories about the effects of curriculum and instruction on student learning.  This talk will present my challenges to, and limited successes with, using evidence-based "scientific teaching,” also referred to as “practitioner research” (with pre-, mid-, and post-tests), to dislodge and correct student misconceptions about evolution by natural selection and scientific epistemology.  At Widener University, which is a 4-year private primarily undergraduate institution in metropolitan Philadelphia, PA, where I teach the fall freshman course in Evolutionary Ecology Bio161, I have been using standardized assessment instrumentation for the past 7 years (multiple choice and essay based).  For fall 2006, I made an extensive set of course revisions (both content and pedagogy) in response to results of 2000-2005 assessment data.  Data from fall 2006 indicate significant improvements in the effectiveness of my teaching (curriculum and instruction) on student learning of core concepts of ecology and evolution.  However, my efforts were much less successful at displacing students’ misconceptions regarding their scientific epistemology, several misconceptions of which appear to be tenacious constructs to dislodge.  I enjoin that practitioner research methods are highly effective means to test hypotheses about the effectiveness of course revisions on student learning. This project is part of a new “Practitioner Research” initiative of ESA’s Teaching Issues and Experiments in Ecology (TIEE, http://tiee.ecoed.net).