How Should We Measure Student Learning? Five Keys to Comprehensive Assessment | Edutopia
Edutopia on Facebook
Edutopia on Twitter
Edutopia on Google+
Edutopia on Pinterest Follow Me on Pinterest
WHAT WORKS IN EDUCATION The George Lucas Educational Foundation

How Should We Measure Student Learning? Five Keys to Comprehensive Assessment

Stanford professor Linda Darling-Hammond shares how using well-crafted formative and performance assessments, setting meaningful goals, and giving students ownership over the process can powerfully affect teaching and learning. Read a blog about implementing comprehensive assessment in the classroom.
  • Facebook
  • Twitter
  • Pinterest
  • Share

Five Keys to Comprehensive Assessment (Transcript)

Linda: The United States is at a moment where it could really transform its assessment systems. Most of our testing is multiple choice tests, pick one answer out of five, which is something you will never do in applying knowledge in the real world. Our assessments need to evolve to reflect the skills and knowledge that we actually value and that we need schools to teach and our children to learn.

Human beings are naturally learners. We are learning every moment of every day. In school, we have particular things to learn and we know that students learn more effectively when we're clear about what the goals are.

Chinasa: I kind of like, I want like a goal in my head about what to do with information that I get. So if they're going to teach me how to do a one page equation, I want to know what can I do with that?

Erin: I look at the beginning of the year and I say, what are the big ideas that I want students to know in the whole year and what are the major skills I want them to be able to do? And then I backwards map those skills and ideas into little chunks. I try to have a really clear learning outcome every day that's measurable, so at the end of class, students can say, "This is how close I got to meeting the outcome for the day." We use the language, emerging, developing, proficient or advanced. If they know where they are on that rubric, then I don’t have to own that for every student and I think that's really powerful.

Linda: Assessment should occur early and often and throughout the process.

Lisa: I do a lot of formative assessments, so formative assessments meaning, it's not summative, it's not a test at the end of the week or the end of a unit. You're doing a lot of small check ins with the students to see where they're at and to see that they're growing a lot.

Teacher: How do you know that?

Student: Because I read it in the Saturn book.

Teacher: In the Saturn books, so you're making a?

Student: Connection.

Teacher: From text to?

Students: Tips.

Teacher: Very good.

Sheela: They are being steered very gently and very strategically, and the formative assessment gathering process is essential to that. You really have to know what kids know and don't know.

Linda: In a lot of cases, we give students written feedback on the paper that they've written but then we move on to the next thing and we don't give them the opportunity to actually rewrite that paper. And the research evidence is extremely clear that one of the strongest positive influences on achievement occurs when students get formative feedback that they immediately can apply.

Chinasa: I think that's one of the things that the kids don't really like, is doing the reflecting, but I think it's actually one of the things that helps us a lot. I think doing it kind of like helps you see the bigger picture, because I don't think that if I didn't do reflecting, I wouldn't really think about what I'm struggling with. I would just kind of want to move on.

Linda: Summative assessment just simply means an assessment from which we can draw a judgment about whether somebody has, at that moment, learned. A false distinction has cropped up in the United States which seems to suggest that it's okay for external summative assessments to just be multiple choice tests. In other countries, summative assessments that occur periodically are essays and oral examinations and project based assessments, so that what you're able to see about what students can do reflects more of what you actually want them to be able to do in the world beyond school.

Jill: So tonight is exhibit night. Every nine weeks, we have this big celebration of learning and the halls will be filled with parents and children who are all here to celebrate what their children have accomplished in the past nine weeks. We also use exhibits to assess the learning of our students. I like to go up to students and have them explain an exhibit to me so that I understand, did they get the big idea? Did they just do a project or do they really understand what they did, why they did it and what it means in terms of the bigger picture? What we find is that because the learning is so rich and it's so meaningful, that our students do very well on standardized tests.

Linda: Many schools that I've worked with have graduation portfolios where students have to complete projects. They often need to collaborate with others effectively as part of the work that they're going to present. They need to communicate in multiple forms. They need to be critical and creative problem solvers because they're going to run into barriers and obstacles and they have to solve them in order to complete this major project. So the blend of the cognitive and the non-cognitive skills that result often in them defending their work to a panel of outside evaluators really prepares them for the range of abilities they're going to need to have when they leave the boundaries of school. And I think it's possible to value these non-cognitive skills throughout school in the same way we value the academic skills.

Erin: We try to bring in people from the community to watch or judge their presentations and then just do things that's like, in your life, in a job, you might have to make a presentation to your boss about why your proposal should work.

Student: So ninety-one percent of the people I surveyed said that fifteen dollars for one unit is a perfect price for them.

Woman: Really nicely done, good presentation and I really like your product.

Erin: So trying to model assessments after real life experiences, I think is really important. No one in their job, I don't think, has to bubble in a standardized test.

Having a hundred and thirty students means it's really easy to have no idea if your students are learning anything because there's so many of them. That's when in the last couple of years, I've realized, the more that I can get them to take ownership of their learning, then I don't have to own that for every student.

Student: Something that I'm good at is researching in books and interviews and articles and to make sure that the information is accurate.

Linda: Ideally, assessment is primarily for the student and the student should own big parts of the assessment process. That can happen in a variety of ways: one is that they are continually engaged in self assessment and peer assessment, using rubrics.

Student: So writing strengths.

Student: Creating good claims.

Student: And then, areas for growth.

Student: I need to analyze my evidence in depth.

Linda: When kids have that opportunity, they begin to value assessment. Assessment is for me, as a way to get the feedback that allows me to become the person I want to be.

If we do this well, the kind of learner we should produce is someone who is self-initiating and self-motivated, understands the standards internally and is continually driving towards excellence, is continually developing their own learning skills, and is able ultimately to learn on their own, independently and collaboratively with others, in a world that is going to be very unpredictable and in which those learning skills are going to be the most important determinant of success.


Get Video
Embed Code Embed Help

You are welcome to embed this video, download it for personal use, or use it in a presentation for a conference, class, workshop, or free online course, so long as a prominent credit or link back to Edutopia is included. If you'd like more detailed information about Edutopia's allowed usages, please see the Licenses section of our Terms of Use.

  • Web Video Producer: Christian Amundson
  • Editor: Daniel Jarvis
  • Associate Producer: Douglas Keely
  • Camera: Christian Amundson, Zachary Fink, Mario Furloni, Josh Gary, Daniel Jarvis, Gabriel Miller
  • Sound: Douglas Keely
  • Graphics: Cait Camarata, Douglas Keely
  • Web Video Strategy Coordinator: Keyana Stevens
  • Senior Manager of Video: Amy Erin Borovoy

Special thanks to ConnectEd, Envision Education, Expeditionary Schools, and Grant Wiggins.

Learn more about the schools featured in this video: Bayview Elementary School, Edwards Middle School, Impact Academy of Arts and Technology, Manor New Technology High School, Marin School of Environmental Leadership, Normal Park Museum Magnet School, Sammamish High School, World of Inquiry School No. 58


Five Keys Video Series

See Edutopia's core strategies in action with our Five Keys video series. Take a deeper look at each strategy as we share the nuts and bolts of program implementation, give voice to examples from schools around the country, and illuminate the research behind the practices.



Assessment is at the heart of education: Teachers and parents use test scores to gauge a student's academic strengths and weaknesses, communities rely on these scores to judge the quality of their educational system, and state and federal lawmakers use these same metrics to determine whether public schools are up to scratch.

Testing forms the bedrock of educational assessment and represents a commitment to high academic standards and school accountability. You can't know where you're going unless you know where you are. But when the financial and emotional stakes associated with standardized tests are disproportionately high, this laudable goal gets distorted. Teachers begin teaching to the test simply to raise scores, often at the expense of more meaningful learning activities. And when the tests are too narrow a measure or aren't properly aligned to standards, they provide little concrete information that teachers and schools can use to improve teaching and learning for individual students.

Twenty-First-Century Assessment

The demands of the today's world require students learn many skills. A knowledge-based, highly technological economy requires that students master higher-order thinking skills and that they are able to see the relationships among seemingly diverse concepts. These abilities -- recall, analysis, comparison, inference, and evaluation -- will be the skills of a literate twenty-first-century citizen. And they are the kinds of skills that aren't measured by our current high-stakes tests.

In addition, skills such as teamwork, collaboration, and moral character -- traits that aren't measured in a typical standardized tests -- are increasingly important. Businesses are always looking for employees with people skills and the ability to get along well with coworkers.

Multiple Forms of Assessment

We know that the typical multiple-choice and short-answer tests aren't the only way, or necessarily the best way, to gauge a student's knowledge and abilities. Many states are incorporating performance-based assessments into their standardized tests or adding assessment vehicles such as student portfolios and presentations as additional measures of student understanding.

These rigorous, multiple forms of assessment require students to apply what they're learning to real world tasks. These include standards-based projects and assignments that require students to apply their knowledge and skills, such as designing a building or investigating the water quality of a nearby pond; clearly defined rubrics (or criteria) to facilitate a fair and consistent evaluation of student work; and opportunities for students to benefit from the feedback of teachers, peers, and outside experts.

With these formative and summative types of assessment come the ability to give students immediate feedback. They also allow a teacher to immediately intervene, to change course when assessments show that a particular lesson or strategy isn't working for a student, or to offer new challenges for students who've mastered a concept or skill. Return to our Assessment page to learn more.

More Edutopia Resources for Comprehensive Assessment

Comprehensive Assessment Overview

Comments (54)Sign in or register to postSubscribe to comments via RSS

JM Ivler's picture
Anonymous (not verified)

As one of the two authors of an issued patent on educational computer adaptive diagnostics I believe that all educational testing today is flawed (of course we are building a product based on our patent to overcome those flaws :-) ).

First, all testing is binary. It is all true/false. Even multiple choice is true/false with each of the distractors a false since they are only used for scoring the "test" and not for immediate feedback to determine where to question next based on the distractor representing a known misapplication of a primitive concept (the core of our patent).

As long as testing is based on a binary relationship between the student and the answer to the question, testing will remain a flawed method of uncovering what a student does and doesn't know.

Just-in-time manufacturing is about getting just what the worker needs into their hands at just the time they need it in order to ensure that they can use it to do their job. Diagnostics sould do the same thing. It defines the limits of the students knowledge and delivers the right tools to the student to learn the things that they need to know at just the time that they are ready to understand those concepts.

If we look at education models that work best we see that directed education using a socratic model is the one that works best. The problem is how do we do that in a classroom environment. Yes, it works well when doing it one-to-one in a tutor setting, but how do we leverage that into a process that can be used in a classroom where there are thirty or more students.

This become more problematic when you consider that the current teaching models are all based on learning segments where the items to be learned are chopped up into one to three week modules of teach-drill-test and then no mater what the result, move on to the next segment. For instance, in a fourth grade class a set of students may be working with fractions using "time" math (understanding the 1/4 hour, etc) for two weeks and then after they test through that they move on to learning about decimal math doing a teaching segment on money. Later in the year they may be introduced to fractions again, but it is not related to time math, and there is no reconnection to the earlier learning segments. So, they are introduced to things like numerator and denominator while not having earlier relevant information included to show how time math and fraction math are similar concepts that are integrated.

This focus on segmented learning in a classroom environment fails to build on the students past knowledge by delivering the tools and materials in a just in time manner. In this way learning is chopped up and there is little building on past exposure to concepts while expanding on new concepts and showing the relationships between them.

Using diagnostics instead of testing allows a process to be developed where a student is presented concepts, and by selecting incorrect answers they expose misunderstanding of primitive components that allows for immediate remediation. So, instead of the segment of teach-drill-test the process becomes a cycle of diagnose-learn-drill. Where the next diagnose moves the student to the next learn process and they can build on the last learned information to expand their knowledge as they build it.

At this point I should state that this method works exceptionally well in any structured learning where the prior information is used to move on. In other words, it works wonderfully in math. If the subject matter is tree-based then this model is the most functional model of teaching. Unfortunately in science for instance the model is more of a coil. There is no need to understand the first law of thermodynamics in order to understand the third. In English it is even more complex as the core to understanding there is usage and a complete understanding of the exceptions to every rule rather than a detailed understanding of the rule itself (think of a scatter shot from a shotgun rather than either a tree or coil).

JM Ivler is NOT a teacher. He is one of the principles of EduCAD Learning Solutions and co-author of a patent on adaptive diagnostics. EduCAD Learning Solutions has completed a years beta testing on their socratic tutor product (a product designed for learning math based on the patent principles) with over 1000 students and is preparing the product for public launch in 2009.

Andrea's picture
Anonymous (not verified)

Although testing can be stressful and demeaning for some people, it can also be a useful tool for teachers to find areas of weakness that must be enhanced. I am in full agreement that testing that simply pigeon-holes kids into the haves and have-nots of intelligence cannot help educators educate. However, a testing system that helps show progress, while still helping us see areas of weakness, would be a great tool for helping students become achievers in academic areas. I do not believe that we need to test in every intelligence area: sports, art, music, etc. I do believe that a basic education in core content is necessary for everyone, regardless of your likes and dislikes, and that everyone can learn to one degree or another. Not everyone is going to be a super-great math student, but everyone can learn basic math and mathematical concepts that apply to almost every area of life. Not every student needs to be able to write the great American novel, but every student needs to communicate effectively with others in both written and spoken language. When we become less focused on punishing schools and more focused on educating our students, we'll see testing as the developmental tool that it should be, not the nemesis of all.

Sarah's picture
Anonymous (not verified)

Hi Carolyne,
Even though standardized tests seem as though they will be here forever, I also do not fully agree with them. I teach 5th grade Special Ed. to students who have mild-moderate learning disabilities. As a Spec. Edu. teacher, part of my job is to accommodate and modify their work so these students are successful. However, the Massachusetts MCAS, has very few accommodations for Spec. Ed. students. My students take the MCAS in every subject. How do you think they feel when they are expected to take the same test as every other fifth grader? My students usually get very frustrated during the test because the test is too difficult for them. I don't understand why I am supposed to assist these children every day at school, but then they are too receive no assistance for a test that they need to pass in order to graduate?! - Very confusing.

Patrick Edwin Moran's picture
Anonymous (not verified)

Our whole educational system suffers because it gives teachers a kind of conflicted identity. The teacher interacts with students, but the teacher is not intended to take the child as a client. A boxing coach should not be the agent of the fight promoter of the gamblers who take bets on boxing matches. Boxers need coaches who are in their corners. But students do not get teachers who can have such undivided loyalty. Teachers are hired by boards of education or analogous groups that have social goals at the center of their mission, and students are put in the care of schools and their teachers by parents who have their own goals in mind. What would be the best for the student is most often not discussed from the student's point of view. Students may not conceive of themselves as individuals who ought to have a say in what they are to learn. Parents may not always honor the students' desires for certain kind of learning. Boards of education may have their own ideas about what students ought to be taught, but also have to keep a wary eye out for parental reactions and governmental imperatives.

Grades frequently function as extrinsic rewards and interfere with intrinsic rewards, The result all too often is that students care about the grades alone, and may learn to associate potentially rewarding course content with tedium and anxiety.

Grades, in most systems, also have the unintended consequence of putting students in an adversarial relationship to those who teach them. Students see teachers as the ones who construct the testing instruments that put them at risk of punitive responses from home, roadblocks to eventually gaining employment, and even the draft. It would be better for the student-teacher relationship if testing were done by an independent figure, and that it be made clear that if a student went to the teacher for a class the teacher's only function would be to prepare the student to pass the examinations.

Grades can perform many functions: They can used as be a shorthand way of informing potential employers, universities, etc. of the level of competence of a student in one area. They can be a summary way of indicating whether a student is ready for advancement along some academic path. They can taken as a measure of intelligence. They can used as a measure of effort. They can function as rewards and as punishments -- even for things that did not even occur in the classroom. None of the above grades is really intended for, or particularly useful to, the student. The student stands in judgment, and the teacher declares the student's worth.

The "rate of progress" measure that grades can give, however, is directly useful to the student, although I have found it difficult to introducing this grading function to students who have been conditioned to a decade or more of the judgmental kind of grading. It would be better for students if they could learn that weekly quizzes are useful indicators of how much progress the student has recently made toward eventually gaining the competency that will allow passage of cumulative exams (midterms, finals, etc.). Students can also be helped to maintain a check-list of course objectives that they will have to meet if they are to have a good shot at the problems to be posed by the external examiner.

In the absence of a real external examiner, the individual teacher may need to approximate this kind of testing instrument as well as possible. Students need to understand what and why this is being done, and they need to understand that high grades lead future employers, etc., to have high expectations that would be dashed if the testing instruments were not accurate.

One way of preparing students to accept this point of view is to make midterms and final examinations from previous years available to them early on in the course. It would be even better if the exams could be shared back and forth among teachers and classes in similar classes. Let students know that these are the expectations of the field, not the whims of a cranky teacher.

A further advantage follows from a well constructed series of semester exams -- teachers are then able to see from class averages where there may systematic weak points in one's teaching, and whether outcomes are improving year by year, are stuck at one point (which would be wonderful if the class average happened to be 95%), or are declining.

A classic on the subject of testing is:

    Preparing Instructional Objectives: A Critical Tool in the Development of Effective Instruction

by Robert F. Mager

Jason Sparks's picture
Anonymous (not verified)

Robert, isn't your recommendation parallel to preparing for the standardized tests such as SAT, GMAT, MCAT, etc. Consider the MCAT (Medical Colleges Assessment Test) market. There's an independent testing authority - the American Association of Medical Colleges (AAMC). Students attend classes through Kaplan, Princeton Review, etc. Others pay for private tutors. Many read books, attend study groups and leverage the Internet using social media and online content.

Robert Siegel's picture
Anonymous (not verified)

I agree with the Edutopia authors that there are many forms of assessment. Unfortunately, the word has received such bad press and been so misinterpreted that its connotation is negative. As in all good things, a balance is required. As I indicate in my Master's level course on "Using Assessment to Improve Learning" at Oregon State University, student motivation for learning is based a great deal on their motivation and encouragement for learning. This requires "ownership". When we move more towards assessment FOR learning instead of assessment OF learning, we begin to see the great power of the concept. There are many strategies to release this power, and master K-12 teachers know how this works. It is critical to employ these strategies as early on in a child's formal education as possible so that they will not feel "jaded" or have a prejudicial view about assessment. This is why I work mostly with K-8 teachers and administrators.
Robert Siegel, EdM. M.A.

Robert Siegel's picture
Anonymous (not verified)

I agree with the Edutopia authors that there are many forms of assessment. Unfortunately, the word has received such bad press and been so misinterpreted that its connotation is negative. As in all good things, a balance is required. As I indicate in my Master's level course on "Using Assessment to Improve Learning" at Oregon State University, student motivation for learning is based a great deal on their motivation and encouragement for learning. This requires "ownership". When we move more towards assessment FOR learning instead of assessment OF learning, we begin to see the great power of the concept. There are many strategies to release this power, and master K-12 teachers know how this works. It is critical to employ these strategies as early on in a child's formal education as possible so that they will not feel "jaded" or have a prejudicial view about assessment. This is why I work mostly with K-8 teachers and administrators.
Robert Siegel, EdM. M.A.

Paige Salane's picture

I strongly agree with project-based learning. I think it is a great way for students to think critically and creativity about a topic. I also think it allows a studnent to freely explore a topic the way they wish. A student will feel good about his/her work and will develop a better self-concept. Assessment is not the answer!

Paige Salane's picture

Project-based learning is a great way for a student to explore a topic on their own. It also gives students the opportunities to critically and creatively think about a topic. Assessment is not the answer!

Sharon C.'s picture

I agree with the article and the videos on the point that standardized testing is not the proper form of assessment. It can be a tool to gauge how students are doing but it does not give the whole, complete version of all that a child can do. Project-based learning is a great assessment tool that teaches students valuable skills for their future career while also exploring the standards in more depth. If project-based learning is taught more in elementary school, students will learn how to do a variety of skills instead of the "intellectual bulimia approach to learning" with tests. Students will learn that what they are taught in school actually applies to real-life. Therefore, learning will seem more meaningful to students. If a project is incorporated at the end of every high school class, hopefully there will be less people that drop out and do not graduate from high school. Students want learning to be useful and relate to their lives personally.

Sign in to comment. Not a member? Register.