Within the last 10 years, higher education has seen the accelerated development of blended learning (BL), active learning (AL), and STEM education. The presented mixed methods study examined a proposed framework that BL affordances enable or constrain AL, as a representation of constructivist learning theory, and that the relationship between BL affordances and AL influences student persistence in introductory science courses. Graham's (2006) definition of BL along the four dimensions of space, time, humanness, and fidelity was paired with Meyers and Jones' (1993) framework for AL as the combination of basic elements to learning, teaching resources, and learning strategies. The BL-AL relationship was then examined along the lines of Graham, Frederick, Byars-Winston, Hunter, and Handelsman's (2013) four dimensions of student persistence--student motivation, student confidence, science learning, and identification as a scientist. The research design followed Yin's (2014) recommendations for a multiple-case study, in which one introductory physics course and one introductory chemistry course were the units of analysis. Interviewed were two student volunteers from each of the following groups: high-, average-, and low-performing. Content analysis was conducted for course documentations on the course website and the learning management system Canvas(TM). In-class and online observations were evaluated using Eddy, Converse, and Wenderoth's (2015) Practical Observation Rubric To Assess Active Learning (PORTAAL) examining student practice, logic development, accountability, and reduction of apprehension. The researcher developed a survey instrument, Blended Learning for Active Learning (BL4AL), capturing students' perspectives concerning AL through traditional and nontraditional learning methods for their persistence in the sciences. Findings indicated students from all levels had varying views of BL affordances in comparison to the original intention of the instructors. Study Edge(c), a third-party tutoring service, was discovered to be an integral component to the students' experiences of BL for AL and student persistence. The regression model using BL4AL responses was found to be significant in the physics course (F(8, 218) = 7.69, p[less than] 0.01, f[superscript 2] = 0.28) and in the chemistry course (F(8, 293) = 6.84, p[less than]0.01, f[superscript 2] = 0.19), and explained 19% of variance in the former and 13% of variance in the latter. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]