r. Peter Ewell is known for the rigor of his scholarship and his effectiveness as a practitioner and consultant in the field of assessment in higher education. In addition, he is an editor and regular contributor to Assessment Update , a bimonthly newsletter published by Jossey-Bass. Early in March, Catherine Kirby, Information Specialist in the Office of Community College Research and Leadership had the pleasure of interviewing him for this edition.

UPDATE: Colleges are at different points in their experience with academic assessment, and community college faculty and administrators are at different points along the continuum of knowledge of academic assessment. To provide an overview of the assessment movement for our diverse audience, could you briefly summarize the iterations the movement has gone through from its contemporary inception in the mid-1980s until now?

Dr. Ewell: A few years ago I wrote a chapter on the history of assessment for a book about accountability edited by Joe Burke of the Rockefeller Institute for Government. In it, I divided the assessment movement into two phases, which I termed, "Round One" and "Round Two." Round One began in the wake of some prominent reform reports on undergraduate education that were issued in the mid-80s, as a result of which a number of states enacted Assessment Mandates. Typical Mandates required all public institutions in the state to implement locally-designed assessment programs and report what they found annually or biennially. Community colleges had to do this very quickly, but for the most part responded well. State level interest in assessment in this form lasted for about five or six years. Around the early 1990s, states began running out of money and devoted their attention to more pressing matters of finance and efficiency. About that time, in what I called Round Two, the accreditation community started to get inv olved in assessment. The feds were putting pressure on accreditors to give greater emphasis to the assessment of student learning in their standards. That immediately provided a second impetus to assessment that has continued ever since.

Those same periods-1985 to about 1992-93 and 1993 until now-can also be distinguished in terms of assessment methods. The early period of assessment largely depended upon surveys and easily available, off-the-shelf kinds of tests. Many community colleges, for instance, concentrated on evaluating the effectiveness of their remediation programs by re-administering placement tests. It was also popular to use tests like the ACT CAAP examination in community college settings, and this approach remains popular for some. But Round Two also saw the development of more authentic kinds of approaches, where the focus is on actual student work. This required faculty to develop scoring guides or rubrics to evaluate actual samples of student work in a more systematic way. During this period, the assessment process was seen less as an "add-on" to the curriculum and became more embedded in it. To say that Rounds One and Two are cleanly divided is probably an overstatement, but it's a nice way of making the point that there has been progress. Your opening question about colleges being at different points in their experiences is a larger issue; there is an enormous variation across individual college experiences, and I think that's important for people to know. There will almost always be somebody out there who is at your college's current stage of development.

UPDATE: I want to talk about both the technique and philosophy of academic assessment. Would you describe the way you conceptualize the integration of assessing teaching effectiveness and the classroom assessment movement within the larger process of outcome assessment?

Dr. Ewell: You use two important phrases in that question-"teaching effectiveness" and "classroom assessment." They need to first be distinguished, and I'll talk about both. Community colleges, like any other institutions of higher education, have faculty evaluation processes that have probably been in place for a long time. A lot of these are based on end-of-course student questionnaires that gather students' perceptions of their instructors' teaching effectiveness. These were in place long before the assessment movement. They haven't gone away because they aren't used for overall curriculum evaluation and instead are used to help make individual decisions about particular faculty. The classroom assessment movement (that actually began in community colleges, by the way) was based on the work of Patricia Cross and Tom Angelo and immediately proved enormously productive and pervasive, especially in community colleges. Classroom assessment was not about evaluating teaching effectiveness so much as it was intended as a tool for faculty to roughly determine how much students were understanding what was going on in a given class session. That movement had a tremendous amount of general impact on the overall assessment movement because people could gather useful information in a manner that wasn't threatening. Faculty could do it in the privacy of their own classrooms. They could see the complete feedback loop of getting information, making a change, seeing if it got better, and so on. Classroom assessment modeled the assessment process in a way that got a lot of faculty turned onto the idea that "this assessment stuff" could actually be beneficial and might be extended to a wider community! At Parkland College, for example, it was interesting to learn that people were beginning to combine their classroom assessments and talk to one another about what they found. So, the privacy issue was overcome by a curiosity issue. I tell that story to illustrate that that's exactly how the larger assessment process ought to work, too.

The literal answer to your question about how to integrate them is that the two processes should be aligned with each other in just that way; learning goals for each class need to be translated to broader learning goals that cut across the whole course, and these, in turn, need to be mapped into learning goals that the department (or the college as a whole) may have for all of its courses. The best assessment programs demonstrate this kind of seamless alignment, so that you can talk to two faculty members in quite different places and say, "What do you expect in a good piece of writing?" (for example) and get consistent answers. So, in terms of integration, you want to have faculty who are practicing the assessment principles in their own classrooms every day.

One of things that is particularly revealing of a good assessment program is when faculty design their own classroom tests to assign regular grades (that is, not as part of an assessment program), they think like assessors and build these tests in the same way that larger assessment programs are put together. Creating routine tests and assignments should be a conscious act with explicit goals for learning in mind. That's the theory of how all these things ought to fit together. There are some institutions that do it well and others that are just beginning.

UPDATE: The dialogue about teaching and learning that resulted at Parkland took on a life of its own as more faculty became engaged and started caring about it. How can we inspire that level of dialogue about assessment in general?

Dr. Ewell: Dialogue must be done across classrooms-that's the bridging part. And we know that faculty are hesitant at first. It's a situation where if faculty can show what they have done in private, then begin to talk about it and reflect on it without having it become a high stakes exercise, we begin to get some familiarity and some collective ownership. The key word here is "collective." Faculty need to hold themselves collectively responsible for certain learning outcomes. The classroom assessment model was the foot in the door that allowed faculty some space to be able to reflect about teaching and learning.

Another thing the classroom assessment story illustrates is that the best way to build faculty ownership is to demonstrate immediate utility. That goes for any kind of assessment. For example, many institutions that are facing accreditation have to build an elaborate exoskeleton, a structure, for assessment. That takes a long time. It's certainly vital to do that, but what you need in between are several little projects that have a beginning, a middle, and an end-situations where some results are generated that faculty can immediately apply to fix something, whether it's looking at the prerequisite sequences in the math program or why so many students are failing a particular course in Psychology. Small research projects like this get people used to the idea that data matters and that we can actually do something about any problems we uncover.

UPDATE: The growth of online education has changed the complexion of student-teacher interaction as well as prompted a rich discussion about student learning. What are the key issues to consider when assessing student learning outcomes in online or hybrid-delivered courses?

Dr. Ewell: Assessment will almost always come up automatically in those contexts because the innovation has to prove itself. Online and hybrid environments have prompted the use of assessment techniques where they might not otherwise have occurred in the traditional environment. I've been involved in a number of projects centered on technology-delivered instruction in both community colleges and in 4-year colleges. In virtually all of them, the assessment approach is basically an experimental design where one must ask, "Is this new approach to instructional delivery as good as or better than traditional one?" That has pressed assessment technique more generally at the institutions where these investigations took place. It has provided a spur for faculty to get familiar with assessment techniques that might not otherwise have occurred.

In terms of key issues, one of the things we're discovering is that learner style makes a lot of difference in these new environments. This means that it's usually a good idea to assess learning style when doing an assessment of online environment because you want to be able to break down results by people with different learning styles. The online environment is one where a lot of individual motivation is often necessary. Picture a learner sitting alone in front of a computer terminal; that's not always a very socially supportive environment. That person has to be self-motivated to get it done. Fortunately for community colleges, adult students tend to be that way. But unfortunately for community colleges, purely on-line approaches don't work nearly as well for younger students who are in academic difficulty or in developmental courses. We've had limited success in terms of online work for less motivated students because they tend to put off the work until a later point and get overwhelmed when they return to it. So the main, extra consideration is the learning style issue. Otherwise, approaches to assessment are the same in the two environments.

UPDATE: The assessment for accountability mentality is replete in the literature. Please describe how you see the NCLB legislation affecting academic assessment at postsecondary levels?

Dr. Ewell: I think it's had a big impact on external stakeholders' perceptions of what assessment is supposed to be about. Back to Rounds One and Two: we've been through this before. The initial flurry of state interest in assessment (back in the 1980s) was very much stimulated by the "Nation At Risk" report about the terrible state of K-12 education, which came out in 1983. We're going through the same cycle again. If you ask a state legislator what ought to be done at the postsecondary level, he or she might reply, "Well, if it is good for K-12, we ought to do it in postsecondary education as well." Clearly, this stance determines the terms of engagement with respect to where external stakeholders want to start the assessment conversation.

Those of us in higher education, particularly faculty, also tend to look at NCLB with a certain amount of skepticism-a great deal of skepticism in many cases. But NCLB is interesting because it holds up high standards, and that's a good thing. Achievable standards are something that we ought to be attempting to emulate. But NCLB also has embedded in it a punishment for non-performance-the consequentiality part of it. Poor performance is always accompanied by getting dinged. And a lot of people only concentrate on the standardized testing part of NCLB. The way to get long lasting assessment going in an institution is to direct resources to fix the problems that assessment detects, not punish the people who have the problem. It is unfortunate that the incentive structures built into NCLB are precisely the ones that can get in the way of healthy assessment at the institutional level.

UPDATE: A large amount of literature on academic assessment is devoted to advancing the art of assessing general education. In your writing you state that the assessment of general education must be integrally linked to the major. Please elaborate.

Dr. Ewell: I guess that's kind of a funny statement to a community college audience because there aren't any majors in the common sense of the word. But there are vocational programs and there are areas of emphasis in the general education (gen ed) track. The larger implication of what I was saying there is that most students only come to know what many of the abilities fostered by general education are all about when they practice them in context. For example, critical thinking means something quite different if a student is taking a chemistry class or a literature class, or if the student is in an auto mechanics program. "Generic" abilities like these manifest themselves differently. Students in any context, and often particularly in a community college setting, have difficulty with general education because there is no clear response to their frequently-posed question, "Why do I have to take this course?" It's much better to have the general abilities that you're trying to foster through gen ed courses embedde d in a context where they mean something to the student. What you really want a student to know and be able to do at the end of a college program have to be manifest not only in general terms but also in some very specific performance contexts. The two go hand in hand.

UPDATE: Career and technical education ( CTE ) faculty have an advantage with respect to outcome measures because of certification and board exams that require specific accountability measures as well as having input from active advisory committees. What could general education faculty learn from CTE faculty, and why doesn't this transfer of knowledge occur more frequently?

Dr. Ewell: My first response, semi-facetious but certainly real, is they can learn that it can be done. There is a track record of being able to do assessment that has already been established at every community college-usually in its vocational and technical programs. The response that the gen ed faculty usually give is, "Well, it's harder for us." In most cases, it is a bit easier in a profession or vocational setting to specify outcomes in competency terms: to say, for example, "you have to put this carburetor on right" or "you need to draw blood in this fashion" or whatever the specific competency is. It really is harder for faculty in, for example, literature, to be able to do that. The point is that faculty can watch what the vocational people are doing and perhaps draw parallels to their own situations.

Career and vocational faculty construct assignments that essentially require a performance. In many cases, arts and sciences faculty do that too, but they don't think of what students do as a performance. But a piece of writing is in this sense a performance, and the question that should follow is, "How can we assess it as a performance?" An interesting example I always pull out when I'm at a campus and this vocational/general education split develops is to bring the fine arts faculty into the conversation: people who may teach dance, or performance, or theatre. What they look for really is performance, but quality rests on expert judgment. Arts and Sciences generally don't have formal "assessments" in place to determine quality, but they are very familiar with juried shows or auditions-that kind of thing. It usually doesn't take much to document the thinking of experts who made such determinations, and other Arts and Sciences faculty can pick up the technique. My own background is in political science so, using myself as an example, I would say, "What elements of an assignment could I construct as a performance (a role play in a scenario-based setting) where an issue comes up and the task is to write a policy brief for a policymaker about it?" That's an equivalent of the clinical settings that health professions faculty have to work with or the portfolio that an art student has to put together. It is down at the assignment level that faculty can learn how to do this.

Now, why doesn't this transfer of knowledge occur more frequently at community colleges? Well, first of all, faculty from these two areas don't talk to each other very often. I've seen some good attempts to remedy this situation through assessment fairs that allow sharing of best practices, or through assessment committees, where people drawn from different programs can get to know each other's practice. I think that both sides have something to learn from the other, and that's becoming recognized. Lurking underneath all of this, though, is a distinction between education and training which some people wrongly advance. This is usually couched as, "You're [ CTE faculty] just doing training; we're [gen ed faculty] doing education." I believe that this is a false distinction and a misperception on the part of the gen ed faculty. But I think there are misperceptions that go the other way, too. For example, CTE faculty often discount the need to instill in students real depth of understanding in the practice disciplines, of knowing the "why" behind the "how."

UPDATE: In constructing assessment with the general education curriculum, how does one go about defining concrete tasks or performance measures?

Dr. Ewell: When approaching this topic I constantly want to return the attention of my colleagues in general education to their own assignments and to ask themselves, "What am I really asking students to do?" One exercise I love to undertake with faculty is to ask them to answer their own test questions. It's surprising how few have actually done so, and then asked themselves, "If I think I provided a good answer, exactly what's good about it?" That's often the starting point for developing a good rubric.

UPDATE: In career courses, technology or advances in practice often drive changes in not only content of what is taught but also the ways learning outcomes are measured. Is there a corollary in gen ed courses?

Dr. Ewell: I think there is, and I think there has been a lot of progress here that can be attributed to both technology and research that ought to profoundly affect what's going on in assessing general education. First of all, there has been a real revolution in cognitive science and as a result, we know a lot more about the way people "make meaning" out of educational experiences. That's affecting the way curricula are designed to include much more collaboration and active learning, much more problem-based learning and service learning. All of those things are just as important in general education as they are in the practice disciplines. They certainly affect assessment in the sense that we now know much more deeply what we're looking for in terms of the meaning-making process, which leads to more authentic ways of thinking about creating scenario- or case study- based methods that can isolate the many misunderstandings that students may have in trying to grasp a concept in, say, physics or about another cultu re. There have also been advances in how to assess team-based projects. We would not have seen any of these a decade ago.

Technology has also changed the way assessment proceeds in some of the general education fields. For example, take the field of writing, where technology now is allowing much more reliable machine-graded answers-ETS now machine scores Graduate Record Examinations, for instance. Similarly, technology provides us much better ways to build online scenarios that can be used to assess critical thinking skills. So in general education, change is not so much about the historical content of key areas of knowledge or ability. It's a little bit different in vocational fields where you may have a completely different way of doing a particular job. The "job" doesn't change that much in general education, but we sure know a whole lot more about what works best in preparing people to do it.

UPDATE: How do the course-taking habits of many community college students, especially those not enrolled in career programs, affect the ability to measure outcomes of general education?

Dr. Ewell: You pose a very important and difficult question because a lot of community college students don't follow the curriculum we design. Instead, they tend to take courses in different orders; they violate prerequisite sequences. Such behaviors present an enormous challenge for assessment. In fact, many of them bring in coursework from other places. This affects the ability to measure outcomes because it is difficult or impossible to definitively establish what the baseline conditions of these students are. The best answer-and I can't emphasize this enough for community colleges audiences-is to undertake sound longitudinal databases that are capable of tracking multiple student paths through the curriculum. These paths provide the "stimulus conditions" that must be associated with outcome measures to make sense of what's happening. If a person performs a certain way on an outcome measure, we want to know what courses he or she has been exposed to, in what order, and under what kinds of conditions. That can have an enormous impact on our understanding of what's really going on.

One great example is in remedial courses, particularly where the timing of the remediation may be crucial to actually being able to master the skills. I remember working with a consortium of community colleges in Texas a few years back and finding that if a student was enrolled in a remedial course in the Fall on an initial placement and entered a math course in the following Spring, the outcomes were much better than in the corresponding Spring-Fall sequence. If there is a whole summer between the basic skills course and the next application, the student tends to forget a lot of content. It's tremendously important to have solid institutional research capabilities at the institution that can tie assessment results to coursetaking behavior and make sense of it.

UPDATE: Can you describe some promising practices you've seen related to the assessment of general education?

Dr. Ewell: The most promising are task-based, scenario-based kinds of assignments that are embedded right into general education courses. Once these assignments are created, they are not an individual faculty member's "property." Colleges using this approach have defined a couple of places in the curriculum where they can collectively develop assignments designed to do "double duty." Faculty get to grade their own students for the record, but then they can rate, using collective rubrics and table readings, samples of these same assignments for more general purposes-for example, to asses critical thinking or problem solving ability. These practices are anchored seamlessly in the curriculum and students don't know that they're being "assessed," so there's no impact on performance and no need for special motivation or anything of that sort. A good example is at Johnson County Community College where key assignments are anchored to general education abilities, pulled on into a cross sectional portfolio, and looked at collectively by the faculty. Another best practice is use of the Community College Survey of Student Engagement (CCSSE) that was modeled on the National Survey of Student Engagement (NSSE). CCSSE is only just getting started and had its first major administration a year and a half ago. But the survey shows all the same signs that the NSSE did of being able to generate a great deal of good faculty conversation about teaching and learning. Again, you need stimulus data as well as the outcomes data to move forward, and CCSSE is another way of collecting this kind of information.

UPDATE: Based on your experience and perspective, describe the ideal makeup of an assessment committee?

Dr. Ewell: Assessment committees tend to be of two kinds. They're either governance committees so they're built into the academic senate or something like that, or they're administrative committees that are advisory to the Cabinet or Chief Academic Officer. I don't have a preference either way, but one thing I don't think is terribly effective is rigidly following a kind of representation model which demands that we have "one of these" and "one of those"-all the kinds of things that typically arise when a governance committee gets involved. The best makeup for a committee, whatever its structure, is a collection of people who are respected by their colleagues. They tend to be faculty and staff who have been at the college a while and who can serve as ambassadors to go back to their own departments and units, let them know what's going on, and build some enthusiasm for the process. A lot of places try to do it as though it were an exercise in pure organizational development instead of emphasizing people skills, a nd that is a mistake. You want the kinds of people who have enough enthusiasm to move the idea of assessment forward, together with a little bit of technical knowledge. It's better to have some people on the committee who know something about statistics and a bit about some of the techniques you'll be dealing with. It's also good to have a couple of skeptics. Invite them in recognizing they are not completely on board and listen carefully to their input. It's a tall order to get all those people but that's the ideal. It's also important to have a succession plan. One thing that happens a lot on assessment committees is that the stalwarts burn out. You have to have some way of being able to renew the committee's membership so people are only there for four or five years at a time, and that they rotate on and off.

UPDATE: Who should the assessment committee report to?

Dr. Ewell: That's a much-debated question, and I don't think there is a single answer to it. It ought to, first of all, have as its principal administrative contact the person who is most responsible for the curriculum. That could be the Chief Academic Officer, or whatever title handles that function. Whether that person should actually be on the committee, or even chair it, is another much-debated question. But, the committee should have a clear reference to the academic affairs side of things. On the governance side, it should report to, or be closely linked with, the authority that approves courses and curriculum. Again, governance arrangements vary between colleges. There is an alternative architecture that seems to work just as well, where the assessment committee serves as the evaluation arm of the strategic planning committee. When the planning committee goes through its annual cycle, assessment is included and takes a look at the effectiveness of the things that were previously planned and implemented. There are a lot of ways to do it and I don't think there is one magic bullet.

UPDATE: How can colleges involve more faculty and build faculty enthusiasm for the assessment process?

Dr. Ewell: Let me re-emphasize the fact that faculty get involved with assessment when they see it as a way to proved an answer to a problem that they actually have. And the problem that they typically have is not something like "the state legislature" or "The Higher Learning Commission." The best approaches to building genuine faculty involvement are the ones that get into the curriculum by raising the question, "What's a particular teaching/learning problem that faculty say they've got? And can we use assessment to build some lines of inquiry that will help them address this problem they think they have?" I've mentioned the topic of prerequisite sequences earlier, but I'll do it again because I find it tremendously powerful in engaging faculty. This is because faculty don't like re-teaching things they thought students were supposed to have learned in an earlier class. That's a common problem. Instead of pointing fingers at other faculty and saying, "You didn't do your job", such situations should be treated a s a collective problem for everybody to work on. Faculty can design a little study to find out exactly why this is happening, then follow up with specially-designed assignments in the earlier course that will test the skills necessary for the later course in the context that students will actually see them. Everyone wins!

Another thing that I wish more administrators understood (actually, I think they do instinctively, but it's hard to act on) is the degree to which the 'gotchya' mentality is the first thing that faculty bring to assessment. The more administrators can reward departments for discovering and sharing a negative finding, and follow that with a plan for how to deal with it, the more they are going to build ownership of the process. What faculty are afraid of is that when a weakness is revealed, something bad is going to follow. If instead, administrators or the committee can, in the first stages of one of these investigations say, "Thank you for helping us learn that we have a problem," and "We have to devote some resources to doing something about it," they create a completely different tone. Language is a big problem too. At bottom, assessment is about inquiry, and faculty respond to the word "inquiry" the word "scholarship." They don't tend to respond as well to words like "standards" or "criteria." The more assessment is seen as a process of scholarly inquiry, the better off you are.

UPDATE: How far are we from the day when, at most institutions, we can shift the focus of assessment efforts from development or process-related issues to using assessment for improvement?

Dr. Ewell: I think we're already there in some places-or certainly very close. We're going to constantly be in the tool development mode, which is a process related effort. The real progress is marked in moving from small scale to large scale. Accreditation has gotten us into this "process improvement" kind of language and that's a good thing. It takes a long time to develop a full-blown assessment process and start using the results. I think it's much more seamless than linear development. The best programs I know started with a few small things, built them up, and gradually spiraled into more utilization of results in a wider and wider arenas. One of the complicating factors is the need to report to external audiences, and that's the accreditation conversation because they want to see assessment processes developing or in place. Processes are what accreditors use to judge whether or not an assessment program is real, is implemented well, and so on. As long as accreditors are focused largely on process it's hard for institutions to get beyond process. They're moving through that phase though, and more and more you hear accreditation conversations focused on use of results. When I advise campuses about the assessment component of accreditation I'll say, "Include some stories, as many as you can, as sidebars in your self study that can illustrate how you fixed something or you discovered a particular problem and did something about it." That's that cycle that accreditors always look for.

UPDATE: Can you describe what academic assessment looks like in a community college where assessment is seen as a natural part of program and institutional improvement and not seen as a requirement of external accrediting agencies?

Dr. Ewell: What you describe is an ideal condition. It exists where there are a few assessment processes (portfolio reviews, CCSSE, or others) that almost everybody on campus knows about. There don't have to be many of them, but they have to be things that virtually every member of the committee knows exist and considers legitimate. Another earmark of campuses that have reached this level is when they have occasions or places when faculty can get together and talk about the results they are getting. An example is a community college in West Virginia that has an annual occasion called "Data Day," where the faculty inservice day is dedicated to looking at assessment data-almost in a party atmosphere. Any of the data can be interpreted a lot of different ways, and faculty have to share those interpretations. The right effect can be achieved by almost any activity where a lot of people get together in small groups around a few key pieces of data that came from the assessment process, then talk about the results and what they are going to do as a result. If an administration is enlightened enough, they will follow up on those conversations with money to respond to the issues and questions that came up. That can build an enormous amount of enthusiasm for the process.

UPDATE: How can colleges begin to use the results of assessment for continuous improvement?

Dr. Ewell: The secret is to start small, start inside the curriculum, put money behind it, and benchmark performance. As we're developing technologies like CCSSE, and increasing the use of ACT's WorkKeys in community colleges, for example, it is important to benchmark performance externally. If faculty can find some publicly-reported information and then compare their own results to them, they can know where they stand and what to do. I know CCSSE going to try to do that in a much more systematic way. Through benchmarking, you can find somebody or someplace that is getting very good results in a particular area and, just as in continuous quality improvement in the corporate world, you can say, "Can we come visit you and see how you do it?" Jeff Seybert at Johnson County Community College is running a benchmarking project like that.

UPDATE: Given that we know that the "assessment as improvement" mentality already exists at some institutions, what must happen for it to occur on a more widespread basis?

Dr. Ewell: The leadership point is particularly important here. You have to have leadership that is ready to stick with this but keep their hands off of the details of it. They have to be visibly behind the process, reinforcing it, and sometimes even packaging their own decisions around assessment results even though their gut may have told them that this was the right decision in the first place. Closing the loop with action is the thing that gets people going; otherwise, people can become cynical.

UPDATE: Most, if not all, of our readers are familiar with your long history of academic work related to assessment. On what specific topics are you currently concentrating your efforts or are most excited about?

Dr. Ewell: I'm working in two very different directions that I'll use as examples. One is "down and in." We're involved at NCHEMS in a couple of projects that are focused on blended forms of instruction using technology. We're designing assessments to try to determine, for example, whether student use of a tablet PC to take notes changes the way they think, and therefore perform. This particular project is not at a community college but it easily could be implemented there. What we're trying to do is create assignments in cooperation with faculty in a sufficiently standardized way that we can count on the results to be valid and reliable to know whether or not the innovation is making a difference. Other things that I'm involved with at the campus level build on that same approach.

At the other end of the scale, I am involved with the Collegiate Learning Assessment (CLA) that the Council on Aid for Education of the Rand Corporation is continuing to develop. This is an authentic, task-based, assessment that poses students with real world problem-solving situations and asks them to address them. Some community colleges in Missouri are administering that examination and using it to benchmark their performance against national standards. We recently participated in a national benchmarking project with that particular instrument, together with the Work Keys, as part of the Pew Forum on College Level Learning. About a dozen Illinois community colleges were a part of that. Some did quite well in recruiting students to participate and some didn't do so well, but it was a great learning experience for everybody. And, I'm still involved in a lot of work with accreditation. For example, I've worked for a long period of time with the Western Association of Colleges and Schools in developing their new collaborative standards-based process that is a much healthier approach than the one it replaced. That gives a range of my current work; we're always involved in something!

UPDATE: Is there anything else you'd like to add?

Dr. Ewell: I guess there are two things. The first is related to our earlier conversation about the unevenness of assessment's development and implementation across campuses. We have colleges that have a good assessment program and then lose it with a change of leadership or some other change of circumstances. That situation emphasizes the culture question yet again. I will occasionally get a call from a place I visited 15 years ago and they don't remember my visit and what might have been learned because many of the people have changed. It emphasizes the fact that assessment is not about processes and organizations-it's about people. Bringing people on board and getting real ownership of the process is something you can never relax about; it's something that constantly has to be worked on.

UPDATE: That's even more important in light of the turnover in administration and faculty in community colleges.

Dr. Ewell: Yes, it's much more turbulent in the community college world. The second thing I would be remiss if I didn't say is that community colleges, despite all of the challenges they face and despite the unevenness in their levels of implementation, tend to be a whole lot better at assessment than are four-year counterparts. That's not uniformly the case, but the fact is that community colleges' main business is student learning, and there is usually a willingness to innovate. This means, in general, they are a healthier place for assessment than most other kinds of institutions.

 Peter Ewell is Vice-President of the National Center of Higher Education Management Systems (NCHEMS) in Boulder, CO. The Center is a private nonprofit organization whose mission is to assist colleges and universities as they improve their management capability. At NCHEMS, his work is devoted to creating longitudinal student databases and other academic management information tools. He has consulted with over 375 colleges and universities and twenty-four state systems of higher education on topics including assessment, program review, enrollment management, and student retention. In addition, Dr.Ewell has authored six books and numerous articles on the topic of improving undergraduate instruction through the assessment of student outcomes. He also serves as a principal partner in the Pew Forum on Undergraduate Learning. Prior to joining NCHEMS, Dr.Ewell was Coordinator for Long-Range Planning at Governors State University. A graduate of Haverford College, Dr. Ewell received his Ph.D. in Political Science from Yale University and was on the faculty of the University of Chicago. He can be reached at peter@nchems.org.

 The interview was conducted by Catherine Kirby, Information Specialist at the Office of Community College Research and Leadership at UIUC. Ms. Kirby's e-mail address is ckirby@uiuc.edu .