Collaborative Testing as a Model for Addressing Equity in Student Success in STEM Classes

Introductory Earth science classes at two-year colleges play a critical role as "gateway courses" for underrepresented student populations into undergraduate STEM programs. Students entering college underprepared in math and science typically receive their only exposure to science at the undergraduate level in introductory courses in the Earth and space sciences. In many colleges a huge disparity exists in these classes between success rates amongst students from groups traditionally represented in the STEM fields and those from underrepresented populations. Closing the equity gap in success in these courses is a major focus of many pilot projects nationally. This concern has also led to the adoption of new teaching and learning practices, based on research in learning, in introductory Earth science pedagogy. Models of teaching practices including greater engagement, active learning approaches, and collaborative learning structures seem to help with student achievement in introductory courses. But, whereas these practices might increase overall student success they have not proven to close the equity gap in achievement. De Anza a two-year college in the San Francisco bay area has a long history in the geology department of incorporating and testing teaching practices developed out of research in learning. Collaborative learning has infused every aspect of our learning approaches in the Earth sciences, including laboratory, fieldwork, and test preparation. Though these approaches seemed to have educational benefit the huge equity gap department-wide persisted between targeted and non-targeted populations. Three years ago collaborative testing models were introduced into our geology and meteorology classes. The mechanism included methods for directly comparing collaborative to individual testing. The net result was that targeted populations including African Americans, Latinos, and Filipinos increased steadily at around 3.5% per year from 66% to 73%. The overall success rates of the non-targeted groups remained between 84% and 86%. Preliminary analysis suggests that for disengaged students in the targeted populations the opportunity to collaborate on a portion of the actual test got them more involved in the collaborative process as it offers immediate tangible return on in-class success.