Higher education is increasingly becoming a prerequisite for success in a technologically sophisticated and rapidly changing world. It helps create an educated citizenry that can strengthen both our economy and democracy. But as a variety of recent studies remind us, its fruits are inequitably distributed between rich and poor – and the gap in attainment is only growing. While the nation has made enormous gains ensuring college access for all students, completion rates are unacceptably low -- just 50 percent graduate within 6 years overall, and even lower for historically under-represented populations. Questions even linger about the instructional value that colleges deliver for the students they do graduate – a point driven home most forcefully in Arum’s and Roksa’s Academically Adrift.
While many factors contribute to student outcomes, one of them is surely the quality of instruction that college students receive. But here’s the problem. The tools and approaches available to measure instructional quality are weak, and improving those tools requires overcoming many obstacles, including the worry that such tools may be used in inappropriate or harmful ways. To appraise the state of the art in measuring instruction in college and also explore possible future initiatives, the William T. Grant Foundation, the Spencer Foundation, and the Bill & Melinda Gates Foundation brought together twenty-two experts on education and the learning sciences in November 2014.
This summary report of the discussion at the convening was prepared to engage a broader audience in the conversation. By capturing the major observations and judgments offered at the convening, it provides a reference point for future discussion and scholarship on the measurement of instruction in higher education.
The report draws on a wealth of knowledge and experience in K-12 education as well as in higher education, and finds a solid foundation on which to build. But there is a great deal more to do. STEM fields have been more heavily studied than the social sciences and the arts and humanities; research in the science of learning proliferates but is not typically applied in a way that demonstrates how different instructional approaches can yield better student learning outcomes; various rubrics and protocols are used – routinely in some instances -- to assess instructional quality but sometimes without as much evidence of their effectiveness as one would like.
The report also articulates a number of crucial design issues that future efforts will need to address early in any work. What ends should instructional measurement serve? (Here participants spoke with one voice: measures of instructional quality should be designed to improve instructional practice and ultimately student learning outcomes and not for high-stakes assessment of institutional performance.) Should efforts be focused on a single discipline or seek to draw lessons from across disciplines, perhaps even leveraging the relative progress being made in the STEM fields to invigorate work in the arts and humanities? Are there obvious starting points, for example, a common taxonomy that distinguishes instructional types or modalities; a review of the research that underpins existing measures and approaches; a focus on instruction in general education?
While answers to these questions are in short supply (this convening was not designed to provide them), insights and pointers to further resources are not. Consequently the report may act as a jumping off point – perhaps even a rallying call for even more deliberate and purposeful discussion. One point of consensus did emerge and it is worth quoting in full from the report’s conclusion. “Better instruction” participants agreed,
"Can make a difference in students’ learning, in their persistence in college, and in their subsequent lives. As more is learned about instruction, more students will benefit. Furthermore, given the number of students who drop out of college, even incremental improvements can make a major difference to the nation.
Understanding instruction and learning how to change it are complex problems. But the knowledge and tools exist or are rapidly being developed to make substantial progress on these problems. The opportunities are great, the participants at the convening agreed, if the means and the resolve can be found to grasp them."