New approaches to instruction are needed in all educational levels in order to develop the skills suited to the twenty-first century (i.e., inquiry, problem solving, innovation, entrepreneurship, technological communication, experimental design, and investigativeness). This research evaluated the outcomes of an approach aiming to develop such skills based on students' assessments of themselves and their peers with regard to investigative projects and course grades. The study is chiefly based on a quantitative paradigm with a multi-method approach. Data were primarily collected using a form to evaluate scientific investigation skills and learning gains; students' project reports and examination papers were the other data sources, which were evaluated both quantitatively and qualitatively. The results revealed a low-to-moderate level of improvement in skills. In addition, the authors discovered that students were overestimating their gains, and that peer-evaluations seemed to function better than self-evaluations.