Doing Assessment As If Learning Matters Most
By Thomas A. Angelo

From the May 1999 AAHE Bulletin

After nearly two decades of uphill struggle, the assessment movement has reached a promising plateau. In general, U.S. higher education has moved beyond unproductive, dualistic debates (remember "four legs good, two legs bad" from Animal Farm?) over whether assessment should focus on accountability or improvement. Today, most faculty and academic administrators have finally, if reluctantly, come to accept that dealing with both is a political and an economic inevitability. Nonetheless, most of us think assessment should be first and foremost about improving student learning and secondarily about determining accountability for the quality of learning produced. In short: Though accountability matters, learning still matters most.

This realignment of opinion hasn’t occurred by chance. Since the mid 1980s, a dedicated and widely dispersed cadre of activists and opinion leaders from higher education associations, regional and professional accrediting agencies, disciplinary societies, and campuses have urged us to use assessment to improve learning quality and productivity. In response, tens of thousands of faculty and administrators on hundreds of campuses have endured speeches, labored in workshops, and conferred at conferences on assessment. Collectively, on committees and task forces, they’ve produced cubic yards of plans, projects, statements, and reports. Thousands have become familiar with, even expert in, assessment. Examples of clever adaptations and creative invention abound. All this effort has been expended despite the fact that involvement in assessment typically counts for little or nothing in pay or in tenure, retention, and promotion decisions. Thus, when most academics "do assessment," personal and professional values motivate them. And the strongest of those intrinsic motivators is undoubtedly the desire to improve student learning.

So the good news is that, over the past two decades, we’ve made impressive progress in assessment. On the other hand, we still don’t have much solid evidence of learning improvement. Why hasn’t so much hard work by so many smart and dedicated people led to better outcomes?

Why Hasn’t Assessment Led to More Learning Improvement?
In his December 1997 AAHE Bulletin article "Organizing for Learning," Peter Ewell provides an insightful response to this question. Commenting on a range of academic change initiatives, assessment included, Ewell argues that our lack of success in improving collegiate learning stems from two common flaws in our change strategies: Initiatives have been implemented without a deep understanding of what "collegiate learning" really means and which strategies are likely to promote it; and initiatives have, for the most part, been attempted piecemeal within and across institutions. Echoing Ewell, I’ll argue that most assessment efforts have resulted in little learning improvement because they have been implemented without a clear vision of what "higher" or "deeper" learning is and without an understanding of how assessment can promote such learning. I’ll also propose that our piecemeal attempts stem partly from a mechanistic, additive model of assessment, which needs to be replaced by a transformative assessment-as-culture-change model if we’re to make real progress.

Three Steps Toward Transformative Assessment
The assessment movement needs a new, more compelling vision to reinspire and reorient our efforts. To steal a phrase from Steve Gilbert of the TLT Group (the Teaching, Learning, and Technology Affiliate of AAHE), we need a "vision worth working toward." Second, we need a different concept of assessment itself, a new mental model. And third, we need research-based guidelines for effective assessment practice that will increase the odds of achieving more productive instruction and more effective learning.

A Vision Worth Working Toward: Assessment for Learning Communities
Our vision of and goals for assessment have led us to less-than-optimal outcomes. For example, we’ve sometimes confused means and ends, doing assessment as if the assessment process matters most, losing ourselves in the technique and method. More often we act as if winning matters most — whether the prize is status, higher funding, or accreditation. Though means and extrinsic ends are important, it’s time to put the highest priority on doing assessment as if learning matters most.

Let me make a comparison to the U.S. space program. NASA, under great political pressure, first focused on winning the space race with the Soviets. When NASA made gains, the agency was rewarded with accolades and bigger appropriations. Along the way, many NASA scientists and engineers became fixated on the scientific and technological aspects of the enterprise — on designing more ambitious, more complex, and more costly equipment and missions. The Hubble Space Telescope and the space station are outcomes of those aims. The overarching aims and vision of space exploration — to discover new worlds and extend our understanding of the universe — were often lost in the shuffle. Thus, even before the Soviets disappeared, most Americans had lost interest in the race and didn’t share NASA’s fascination with the technology. Now NASA is struggling to rebuild public support by focusing more attention on cultivating the intrinsic human interest in discovery — our collective desire "to boldly go where no one has gone before." 

In a widely read and discussed article, Robert Barr and John Tagg developed the thesis that U.S. higher education is in the midst of a historic shift from a teaching-centered to a learning-centered paradigm ("From Teaching to Learning," Change, Nov/Dec 1995). In this emerging paradigm, Barr and Tagg see the primary purpose of colleges and universities as producing learning rather than providing instruction, and traditional teaching as only one of many means of learning production. Drawing inspiration from Barr and Tagg, I predicted in a 1997 article that one major outcome of this paradigm shift would be the transformation of colleges and universities from "teaching factories" into "learning communities" (see "The Campus as Learning Community," AAHE Bulletin, May 1997). By learning communities, I meant carefully designed groupings of students and faculty working intensively and collaboratively toward shared, significant learning goals — often by focusing on themes that cut across several traditional disciplines.

As K. Patricia Cross pointed out in ACPA’s July/August 1998 issue of About Campus, strong support for engaging students in interactive, collaborative learning communities can be found in the research on learning outcomes, on development, and on cognition and motivation. She also noted important pragmatic reasons for creating learning communities, among them workforce training and citizenship education. 

In my view, the learning communities ideal and many of its best current manifestations represent a vision worth working toward, not just for assessment but also for educational change efforts in general. Having the construction of learning communities as a goal is quite different from aiming at incrementally improving our present system. It’s a whole new ball game. If we accept, at least for the moment, creating productive learning communities as an orienting vision, then our concept of assessment must also change to support that vision.

Changing Our Mental Models: Assessment as Culture Transformation
The second reason our assessment efforts have been less successful than desired has to do with our concept of assessment itself. Some view assessment as a mechanistic, technical process — a collection of monitoring and problem-solving devices that can be dropped into or added onto existing academic programs, much as we might connect an antismog device to a car engine. Others, probably conditioned by program evaluation and accreditation experiences, see assessment as a necessary, periodic bother, like a visit to the accountant at tax time. In either case, assessment is seen as something that might result in small changes, usually adding data-collection and reporting processes here and there. At this point, we have enough collective experience to realize that these additive, episodic approaches to assessment rarely work or last.

To improve learning and promote learning communities, we must recognize that successful assessment is not primarily a question of technical skill but rather one of human will. To return to the NASA analogy, all the advanced rocket science in the world is of little use if there is no widely shared interest in exploring the universe. Assessment may not be rocket science, but the same principle holds: Assessment techniques are of little use unless and until local academic cultures value self-examination, reflection, and continuous improvement. In general, already existing assessment techniques and methods are more than sufficient to meet the challenges we face. It’s the ends toward which, and the ways in which, we use those tools that are the problem.

Four Pillars of Transformative Assessment
I don’t believe we can construct learning communities with our students or practice transformative assessment unless we first develop what Peter Senge, in The Fifth Discipline (Doubleday, 1990), calls "personal mastery." Thus, in order to move beyond piecemeal and superficial change and toward transformation, we need to develop a learning community-like culture among the faculty and administrators involved in assessment. Four basic preconditions are key to this collective personal mastery. First, we need to develop shared trust; second, shared visions and goals; and third, shared language and concepts. Fourth, we need to identify research-based guidelines that can orient our assessment efforts toward the goal of creating productive learning communities. (Since all these preconditions need to be developed at more or less the same time, their presentation order is relatively arbitrary.)

Build shared trust: Begin by lowering social and interpersonal barriers to change. Most of us learn little of positive or lasting value from people we don’t trust. To form a productive learning community, the faculty involved in assessment must first come to trust one another. Let me suggest a simple first step. Before turning to the problems, tasks, and issues to be resolved, take time to highlight what participants are doing well and to share successes. Encouraging participants to share examples of successful teaching or assessment practices allows them to present their best face and demonstrates that each is a smart person with ideas to contribute. Whatever the means, the point is to start not with problems and debate but by helping participants feel respected, valued, safe, and in the company of worthy peers. 

Build shared motivation: Collectively identify goals worth working toward and problems worth solving — and consider the likely costs and benefits. Since goals powerfully motivate our behavior, developing a set of shared learning/assessment goals is a logical next step in building a productive learning community once shared trust has been established. Most of us are more productive when we’re working toward clear, personally meaningful, reasonable goals. While students and faculty members typically have goals, they rarely can articulate what these goals are, rarely know how well these goals match their peers’ goals, and rarely focus on learning. Faculty goals tend to focus on what they will teach, rather than what students will learn; student goals often focus on "getting through."

There are many techniques for assessing goals, but the key is to find learning-related goals in common. The "Teaching Goals Inventory," developed by K. Patricia Cross and me, is a (non-copyrighted) quick self-scorable questionnaire to help faculty identify their most important instructional goals. You can find the Inventory in Classroom Assessment Techniques (Jossey-Bass, 1993). An even simpler approach is to ask faculty to list two or three assessment questions they would like to see answered in the coming year, or things they would like to ensure that students learn well before graduating, and then look for common goals across the lists. Whatever the shared goals, in order to be useful they must be clear, specific, linked to a timeframe, feasible, and, most important, significant.

Goals are not always sufficient to motivate us to learn. After all, if the status quo is not problematic, why change? But not all problems provide useful starting points. As Ewell noted, "Maximum learning tends to occur when people are confronted with specific, identifiable problems that they want to solve and that are within their capacity to do so." In any case, it’s critical to connect and frame problems within a larger vision of shared goals so that energies and resources aren’t dissipated in myriad efforts that add up to little or no improvement.

Here’s a three-step thought exercise faculty and students can use to identify promising assessment problems. First, once you have a problem in mind, write down what you think the best solution would be. Second, assuming that were the solution, could the group actually implement it? And third, even if it could be implemented, would the group choose to do so? If the answer to either of the latter questions is "no," it’s probably not a problem worth taking on. 

If the answers are "yes," then it’s time for a cost-benefit analysis — however informal — of the proposed solution.

I recommend trying to "guesstimate" the following types of costs before committing to an assessment problem: costs in human time and effort, costs in financial resources, costs in political capital, and opportunity costs (i.e., what other important problems won’t you be able to tackle if you follow this path?).

Build a shared language: Develop a collective understanding of new concepts (mental models) needed for transformation. Building a shared vision for transformative change requires shared mental models and shared language for describing and manipulating those models. In other words, before we can collaborate productively we must establish what we mean by terms such as learning, community, improvement, productivity, and assessment. Taking this step will allow us to make any implicit conflicts among our mental models explicit so that we can work them out. 

One simple strategy for uncovering different mental models is to ask faculty to define in writing what they mean by one key term, such as assessment. Then collect those responses and discuss them or create a concept map from them, making visually apparent the areas of agreement and difference. You may find that assessment means, variously, standardized testing, student ratings of faculty, grading, institutional research, and time wasted. Rather than arguing for one correct definition, I suggest proposing the adoption of an additional, shared working definition, much like adding another meaning after a word listed in a dictionary. This strategy doesn’t force individuals to change their mental models, something many will resist. Rather, it asks only that they acknowledge differences between their models and the group’s and that they use the group model when collaborating.

Build shared guidelines: Develop a short list of research-based guidelines for using assessment to promote learning. Several lists of guidelines for effective assessment already exist, most notably AAHE’s "Principles of Good Practice for Assessing Student Learning" (posted to AAHE’s website, www.aahe.org). What I’m suggesting here, however, is that individual campuses and programs can benefit from constructing their own specific lists of principles or guidelines — lists that can also serve as the criteria for evaluating their own assessment plans and efforts. As an example, here’s a list of principles I’ve developed, based on my reading of the research on improving learning and development.

10 Guidelines for Assessing As If Learning Matters Most
If learning really matters most, then our assessment practices should help students develop the skills, dispositions, and knowledge needed to: 

• Engage actively — intellectually and emotionally — in their academic work. 
• Set and maintain realistically high, personally meaningful expectations and goals. 
• Provide, receive, and make use of regular, timely, specific feedback. 
• Become explicitly aware of their values, beliefs, preconceptions, and prior learning, and be willing to unlearn when necessary. 
• Work in ways that recognize (and stretch) their present learning styles or preferences and levels of development. 
• Seek and find connections to and real-world applications of what they’re learning. 
• Understand and value the criteria, standards, and methods by which they are assessed and evaluated. 
• Work regularly and productively with academic staff. 
• Work regularly and productively with other students. 
• Invest as much engaged time and high-quality effort as possible in academic work. 

The limits of this article do not allow me to illustrate the guidelines above, but many appropriate examples can be found in the assessment literature. But those examples, and any list of assessment guidelines, will only be useful to the extent that we, the assessment activists, first establish the fundamentals. To achieve transformation in higher learning, we must develop shared trust, a transformative vision of goals worth working toward, and shared language and concepts equal to the challenge. If we plan and conduct our assessment projects at every step as if learning matters most — and not just student learning, but ours as well — then the distance between means and ends will be reduced and our chances of success increased.

Reprinted with permission from the American Association for Higher Education