Why Assessment Is Not the EnemyShare
Any kind of assessment – whether a classroom quiz or the state’s annual academic achievement testing in math and reading – is only worth giving if the time and money used to give it pays for itself in the form of greater student learning.
There can be little doubt that standardized tests are a driving force in American education. From college entrance to kindergarten entrance and everything in-between, it seems that for every problem in education a new assessment has been posed (and derided) as the solution. That said, the mechanism through which testing is supposed to lead to better learning is at best unclear and at worst seriously confused by policymakers and educators alike. A seminal study by Urdan and Paris (1994) found that most teachers possessed negative attitudes toward standardized tests even to the point of many disrupting or downright sabotaging the process. And this was in the 1990s – long before the explosion of testing under No Child Left Behind. Most teachers have either experienced first hand or know someone who has had to watch students sit for hours on end and fill in bubbles only to wait several months – often into the following school year – to receive any results back. And when these results do come back, the information seems at best overly judgmental of students and teachers or at worst completely invalid and a betrayal of everything education stands for. Either way, the results are often filed away quickly and quietly while everyone involved takes a deep breath until the next round of testing.
But why are tests so controversial? What went so wrong that an activity like questioning what students know and are able to is now seen as such heresy? What I propose in this article is not that standardized tests are victims of a vast conspiracy theory, but rather that the larger culture of American education has lost its way regarding the purpose of all forms of assessment and what role they can and should plan in the teaching of students.
Any kind of assessment – whether a classroom quiz or the state’s annual academic achievement testing in math and reading – is only worth giving if the time and money used to give it pays for itself in the form of greater student learning. We in the world of K-12 education do not give tests because we are mean people or because our students like them. In this there is likely little disagreement. Instead, the primary purpose of assessment in schools must be to inform instruction in such a way that students learn more, or at the very least more efficiently, because of how we as educators use the resulting data. For example, if a quiz I give in my math class tells me that 75% of my students already know the content I was going to teach next week, I can modify my plans so that students actually learn something new instead of simply being re-taught old content. In this case a 20min quiz saved students time in class that in turn allowed them to learn new content and skills. Likewise if I find that students do not have the prerequisite skills in order to benefit from the content I was about to teach, then I can back off and review those skills before moving on thus making my teaching and student learning more efficient. This guiding principal of beneficence is the root of one of the greatest issues surrounding standardized tests – many do not end up providing educational benefits for students because their results are not used, are un-interpretable, or do not actually yield information that is helpful to teachers and administrators when making instructional decisions. I will address each of these issues in turn but first, a long-winded golden rule:
If the information yielded from assessment – whether standardized tests or classroom quizzes - will not be used to inform instruction then the assessments have absolutely no reason to exist.
Barriers to Assessment-Based Decision Making
If teachers and administrators will not have time to sit down and review the results of a certain test and how the information can be used to better student learning, then the test itself cannot possibly improve instruction. Simply giving most tests does not in itself yield student learning – though some student assessment can be educational. Similarly, if teachers and administrators are so passionately against the very idea of assessment in education that they refuse to even consider the results, then the district can save its money and the students’ time and not give the tests. Again, any test is only worth giving if the benefits outweigh the costs both in terms of money, student time, teacher time, and any negative side effects to morale, stress, etc. If the benefits (no matter how small) will not be realized, then the cost (no matter how small) cannot be rationalized. Some of this educators and administrators cannot control. For example, in the past standardized tests took a lot of time for students to take; there was no way around this. Now, with tools like the Measures of Academic Progress® (MAP: Northwest Evaluation Association) and the new Smarter Balanced Assessment System’s summative assessment being computerized and adaptive, the tests will take much less time1 – they come at a smaller cost in terms of lost instructional time. This can eliminate one significant barrier to the successful use of assessment to inform instruction. However, the barrier regarding educator willingness to use the results remains.
Even though educators cannot control how long standardized tests take, they can control the benefit side of the equation. Teachers and administrators can control how much benefit students receive from the tests they took and those educators should strive to get students every possible return for their investment in time. Below are some suggestions for how educators can make the most of the assessment data.
Time Alone (or with friends) with the Data
There are only so many hours in a teacher’s day. The lost instructional time required for standardized tests often leaves teachers with little time to consider what the results mean and administrators with little appetite to grant instructional releases for teachers to apply any instructional differentiation. This has resulted in piles of data (sometimes literally) sitting in an office somewhere that no person has ever looked at in terms of what it could mean for instruction. In this case the purpose of the test was simply to be given – not to be used. This is why I argue that assessment and testing cannot end with administration. In fact, James Popham, a major critic of how assessment is sometimes used in education, suggested that assessment and measurement are different concepts in that measuring student learning (through testing) only becomes “assessment” if the data collected were somehow used to inform instruction (2011). If there isn’t time once the results come back for teachers to work in data teams to apply the lessons learned from the testing, then the measurement of student learning can never become assessment – it can never help teachers improve student learning. If this is the case and there simply isn’t time to ever use the data gained to inform instruction, then the test should not be given.
Results and Reports
The next issue is just as common as the last – test results come back and are reported in such a way that even an educated person cannot easily understand. In some cases results are reported to teachers in the form of a single number. What this number means when it comes to how a single content-area teacher should teach his or her class is left ambiguous. Similarly, some reports come back as a single page for every student making classroom-level instructional modifications time-consuming if not impossible. Luckily much of this is changing. With computer-based adaptive tests, results are often reported more quickly (sometimes instantly) and alongside a wealth of resources for interpretation. The barriers standing between test data and the consumer are starting to come down, theoretically making the use of test data easier for the classroom teacher or school administrator. Of course, no data or analysis tools will make any difference if teachers and administrators do not have the time or interest in using them. Some training will be required in how data from assessment can be used to inform instruction if the potential benefits are to be realized.
What Can the Test Tell You?
By far the largest barrier to using standardized assessment information to inform instruction is the degree to which the data yielded are actually useful for a teacher’s instructional planning. For example, a school could given an intelligence test to every kid, but what would that tell me about the poetry lesson I was going to give on Monday? Similarly, many state achievement tests report global scores related to language arts and math but are not detailed enough to show which specific math or language areas a student needs to further develop and which skills have already been mastered. This utility of data is often called diagnostic efficacy – the degree to which the test provides useful information that I as a teacher can use to modify my instruction. Data such as IQ scores or global measures of a student’s language arts mastery provide low diagnostic efficacy whereas tests such as the MAP® provide far greater detail regarding what students know and are able to do. This higher level of diagnostic information makes the data more teacher-friendly in terms of being useful to informing instruction. Now, before I give that poetry lesson, I can check and see if the students in my third-period class have already demonstrated mastery of those skills or if they need more remedial material first.
Principles and Principals
I have personally witnessed schools in which teachers or administrators took the principled stand that standardized achievement tests are not appropriate in education. This has taken the form of teachers who refuse to look at student test scores or even school staff telling parents to refuse to let their children take the state test. I do not at all blame teachers for wanting to spend more time teaching and less time testing. Rather what concerns me is the belief that no useful purpose can come out of assessment – the pervasive idea that knowing what students already know cannot possibly be useful because all kids should be taught the same. The negative sides of assessment and testing have been the focus for so long that many educators at every level have forgotten that there is a primary purpose to all assessment: informing instruction. They are not perfect and they come with many negative side effects, but if the results are not used then we only exacerbate those side effects by removing all potential benefits. There is some level of useful information in nearly every set of standardized test scores.
Do we have any Control?
Of course, the elephant in the room is that the decision of whether or not to administer a test is often made at the state or national level – far beyond the control of the concerned teacher or principal. The Federal No Child Left Behind Act (as well as most state waivers from it) requires certain subjects be tested at certain times leaving schools with little control over what is given and when. In most instances the cost part of the equation has been decided for us. However, teachers can control other tests given in their classrooms and apply the same principles discussed earlier to make sure that every test given is done so to improve instruction. If I am not going to look at the results of my unit two quiz in order to better refine my lesson and my teaching, then I should not have given the quiz in the first place. If my students average a 60% on the quiz but I do not go back and review the missed material, then what was the point in giving the quiz? With regard to standardized tests that are mandated, teachers and administrators can also make sure that the maximum amount of benefit is squeezed from every test so that student time taking these tests was not in vain.
I recognize that this philosophy is overly idealistic – that in the real world assessment and testing are far more about grading and accountability than about student learning and instructional planning. Although grades often do stem from assessment and standardized tests are now used for student growth monitoring and educator effectiveness measures, their primary purpose must come back to student learning. Tests and assessment can be used for many purposes in education with some being reasonable and helpful and others being questionable or disturbing. However, there is nothing disturbing about squeezing every bit of useful data out of every assessments students take. No matter how little data I get from a particular test, it’s more than I had before and every little bit of data can help me better teach the diverse range of learners that are my charge.
Popham, W. J. (2011). Classroom assessment: What teachers need to know (6th ed.). Upper Saddle River, NJ Pearson.
Urdan, T. C., & Paris, S. G. (1994). Teachers’ perceptions of standardized achievement tests. Educational Policy, 8, 137 – 156.
It’s important to note that in some Smarter Balanced States the new test will actually take longer than the old ones. While this seemingly contradicts my point form above, the reason it’s longer is because it is a better / more-valid measure of the higher-level standards it is supposed to assess. It is also getting into application skills whereas past tests struggled to get beyond basic recall.