Assessment | Viewpoint

The Testing Straitjacket

• By Trent Batson
• 07/07/10

How students are tested or evaluated determines how they are taught. But testing within many courses today remains, in essence, the same as always. Therefore, the limits of educational reform are determined to some extent by the current legacy structure of testing and evaluation.

It is true that early adopter faculty on campuses across the country have, in many cases, moved on to alternate methods for student evaluation, but early adopters are not representative of the mainstream, and their methods are therefore usually not transferable, except to other early adopter faculty members.

For successful and pertinent educational reform, testing and evaluation practice must first change to bring it into harmony with the fundamental realities of the knowledge economy and prevailing knowledge tools of today.

What is it about current testing practices that is so anachronistic?

 Legacy testing usually requires students to remember lecture notes and readings, or the results of lab experiments (or field experiences, etc.). Trying to use any source to prompt the memory is usually considered cheating. The tacit belief seems to be that memorization is vital to success in life. But is this true? In an age when Google or Bing is a finger-stroke away, why are we training students to memorize rather than to be quick to use the readily available resources of our time? Isn’t this what we do in real-life work? Not that memorization skills aren’t important but it seems disproportionately emphasized in our legacy testing procedures.

 Testing is almost always about “given,” not discovered information and knowledge. Traditionally, teachers lecture, assign reading in books, conduct some classroom discussion, and then test how well the student remembered this given information. But, is this the way a college graduate would be expected to work after graduation or even succeed in graduate school? Most real-life work for college graduates is solving one problem after another where there may be almost no given information at all: Usually, college graduates would be expected to discover information and knowledge in whatever work they do.

 Testing often does not refer to or allow students to use the work they’ve done leading up to the test. They will certainly use the ideas from their assignments, but are generally prevented from referring to their work during the test.

And so on. Many more examples are available, but these three alone suggest we have problems with our current predominant model of testing, and also suggest that testing as commonly structured today is out of date.

Portfolios

Now, let’s assume a teacher decides to try a portfolio approach in her class instead of using these legacy testing practices.

Students could then upload their work into a portfolio available on the Web. The teacher can then see the student’s work within the portfolio. The students own this work--legally and because they control who sees it, and because they can keep this work after the course is over (unlike work in a course management system which goes away after the course).

If the portfolio is organized well with tags and good search features to help, students can quickly refer to this work (that is, reflect on it) and select some of it to create a Web page presentation of their work for various purposes.

Portfolio methods have been used when only paper was available, but much more so now that portfolios are electronic. But, portfolios require a different approach to teaching and learning, so the course must be changed if portfolio practices are to succeed. Think of the testing example. Think how different the legacy testing paradigm is to a portfolio paradigm:

 Testing, as now practiced in most courses, refers to “teacher knowledge,” depends on memorization, and gives scant recognition to student discovery.

 Portfolio practices include teacher knowledge and student knowledge as the over-all reference material, encourage students to use their collection of evidence as a strong developmental practice, and fully recognize the value of student discovery.

The two (the current legacy testing practices and the portfolio assessment practices)--cannot co-exist since they are opposite in all ways. One cannot both employ a portfolio in an important way and continue to test as always. The portfolio emphasizes the centrality of reflecting on student work, a meta-cognitive skill, while legacy testing emphasizes the centrality of remembering teacher knowledge, a memorization skill requiring little meta-cognitive engagement. Students need to know which is important, and if the test determines the course grade and not the portfolio, then the portfolio will seem superfluous. Either the portfolio or the test then becomes irrelevant in the course, and, up until recently, teachers were more likely to discard the portfolio than the test.

The New Landscape

The Web and education technology will be with us in some form from now on. They are and will be the knowledge tools of our culture. Since, therefore, almost all academic work is and will be digitized and therefore logically can be stored in some way on the Web, we need someplace to put it and manage it over time. That “place” is an electronic portfolio. Why is Facebook so popular? Because it provides portfolio-like capabilities. But, it’s not a personal portfolio.

A car is to the interstate highway system as a portfolio is to the World Wide Web: Both are personal spaces that we control but which allow us to function in either the analog (the car) or the digital (portfolio) world.

Since portfolios will most likely serve this fundamental mobility we need to live and work and learn in this century, the question is probably not whether you will use a portfolio or not, but when you will use a portfolio--and will you be able to drive it?

Portfolios dictate a different approach to evaluation: accumulation of work evidence and reflection on that work. Using reflection as the most basic way that we academics evaluate students is far more appropriate to the way we work in this century than the testing methods of last century. We no longer need to test as we did, but we do need to evaluate using portfolios.

[Photo by Trent Batson]