E. Jean Gubbins
University of Connecticut
You’re having a test tomorrow! When was the last time you heard or actually said this statement? How did you respond? Excited! Panicked! Disinterested! Motivated! Perhaps you need more information, such as: What subject? What type of test? How long will it be? Does it count toward my final grade? Depending on the answers to these questions and others, your motivation or anxiety may increase or decrease. Your past experience with tests may influence your reactions to a great extent. Did the awareness of a test on a certain date help you focus your learning or, at the last minute, get ready for cramming?
You may react to tests in many different ways—as a challenge or a nightmare. In reality, tests are to inform you, your teachers, your parents, your administrators, and your community. Information resulting from tests should guide content and instruction, rather than just something that happens after a specific number of weeks pass by in classrooms around the country.
Through our research at The National Research Center on the Gifted and Talented (NRC/GT), we have learned a lot about how tests are used in classrooms. Of course, our findings and conclusions take on different perspectives, depending on the types and purposes of tests. For this article, a narrow perspective on achievement tests is offered: group assessment consisting of objective, close-ended items focusing on a specific set of content objectives. Essentially, test items are constructed in response to an overarching question: What do you know? There is no attempt to find out how you know this information, how you can possibly demonstrate your knowledge in alternative formats, or how you can apply this content knowledge to similar or novel problems or situations.
There are numerous books to consult about the history and dynamics of testing (cf. Elmore & Rothman, 1999; Lyman, 1998). Some are very technical; others are step-by-step approaches to designing tests that measure students’ content knowledge. Books explore the world of testing as a science and an art. Tests and subsequent test scores grab people’s attention. When someone states test results, he/she sounds authoritative. There is an instant acceptance of the data as truth. Test development is a serious business. Some people embrace tests as an objective measure in response to a basic question: How are we doing? Others view tests as an intrusion on the true meaning of learning that goes beyond mastery of content knowledge. There are probably more viewpoints about testing than books about designing tests.
One viewpoint about tests is from a children’s book. Children’s book authors often capture the meaning behind situations, issues, or problems in such a clear, consistent way because they are writing for and appealing to young people. Their stories and messages don’t escape the adult mind. However, the stories may escape us because of lack of access.
How often do you read children’s books?
frequently sometimes infrequently never
If you selected “never” in response to the test question above, then you may have missed a wonderful interpretation of the impact of tests from one of the wisest children’s authors and unproclaimed philosopher—Dr. Seuss. Knowledge, sentiment, understanding, and celebration come together in Dr. Seuss Hooray for Diffendoofer Day! (Seuss, Prelutsky, & Smith 1998). Read the final section of this book first. It is customary for authors to explain why they wrote the book and then acknowledge people who supported the process and made it possible to share it with others. This information is usually part of the preface. Hooray for Diffendoofer Day! turns the protocol of the book world topsy-turvey. After the story is a section entitled: “How this book came to be.” The basis for this book was a creative idea sketched out by Dr. Seuss’s many musings as he played with words, titles, and drawings of people and places. You have an opportunity to trace his thoughts and ideas and witness the brainstorming process in which he engaged alone in his studio. Years went by and the potential book idea was referred to infrequently. Unfortunately, the book never reached completion during Dr. Seuss’s lifetime. However, the treasure trove of ideas and illustrations did not remain hidden from all. With the support of his former editor and Alfred A. Knopf, Inc., two dedicated professionals completed the book or as the book jacket states—Dr. Seuss with the help of Jack Prelutsky and Lane Smith. The team of Prelutsky & Smith consists of a famous children’s author and illustrator, respectively. They honored Dr. Seuss by completing this book. In a way, they, too, had to “pass the test” of bringing Dr. Seuss’s musings and drawings to life and casting a story with meaning beyond words set in print.
Three short sections of Hooray for Diffendoofer Day! present a view of tests at Diffendoofer School that may ring true in your school:
His name is Mr. Lowe.
He is the very saddest man
That any of us know.
He mumbles, Are they learning
This and that and such and such?
His face is wrinkled as a prune
From worrying so much.
Later in the book Mr. Lowe announces:
Must take a special test,
To see who’s learning such and such—
To see which school’s the best.
If our small school does not do well,
Then it will be torn down,
And you will have to go to school
In dreary Flobbertown.”
Of course, students took the test and they all waited for the results:
Mr. Lowe meandered in.
We’d never seen him smile before,
But now he wore a grin.
He soon began to giggle,
Then his giggle grew by half,
And then it really happened—
Mr. Lowe began to laugh.
“You’ve saved our school!
You’ve saved our school!”
He jubilantly roared.
“We got the very highest score!”
He wrote it on the board.
Obviously, all of the students in Dr. Seuss’s book had perfect scores. Such is the reality of endings in some children’s book.
Returning to the reality of school in 2001, we know that tests can be very useful in determining mastery of curriculum, assessing student progress over time, maintaining a system of accountability, and providing one view of performance. We must know why we are testing students, how we are testing students, and what we are going do with the resulting data.
In the NRC/GT research protocols, we use tests in multiple ways. Depending on the particular study, we may want an “insurance policy.” For example, we know that many gifted and talented students have actually mastered the curriculum planned for their grade level prior to the first day of school. Does that sound impossible? In our study of curriculum compacting (Reis et al., 1993), we found that high ability students (grades 2-6) mastered 40-50% of the traditional classroom material in one or more of the following subjects: mathematics, language arts, science, and social studies. Try it yourself. Ask last year’s elementary teachers to give you the names of the top 3 students in reading, spelling, or mathematics. Select a test that you would normally use at the end of the school year or choose a unit test from your teacher’s manual. Administer the test to the top 3 students and determine the extent to which they know and understand the content.
This curriculum “insurance policy” is the documentation of what the children know. Obviously, we would not want to eliminate or streamline curriculum if the student could not prove mastery of specific grade level objectives. A profile of what children know allows us the luxury of considering what they want to know and, possibly, how they want to learn the new information and skills (Starko, 1986). Students may work with the next level of complexity in a specific content area or use their current content and skill mastery to extend learning across disciplines. For example, one second grade student was particularly adept at poetry writing. She created poems using many styles and formats. Her choice of topics was also wide ranging. She captured the essence of language and enjoyed sharing poems with others. To further the development of her poetry skills, she worked with a local poet. With a mentor, this young person escalated her writing ability as a poet and started working on developing original plays. Her language arts time was adjusted to meet her learning needs. She and her mentor worked together twice a week during language arts. Periodically, this student’s skills were checked with readily available unit tests to ensure that she continued to know, understand, and use grade level and above grade level skills to a high level of accuracy. With these assurances of the mastery of content and skills, the classroom teacher completely supported the elimination of grade level curriculum in language arts on a unit-by-unit basis for this young person.
When we studied the impact of programming for gifted and talented students (grades 2-3), cognitive and affective variables were of interest. In the quantitative study of learning outcomes, Delcourt, Loyd, Cornell, and Goldberg (1994) used achievement tests to look at the cognitive gains of programming using various service delivery models: special class, special school, pull-out program, and within-class program. We administered pre-post, standardized, norm-referenced tests for 2 years in mathematics, reading, science, and social studies to determine growth over time. We considered using tests that were one grade level above the students’ current grade assignment. We experimented with a small group of students and found that out-of-level testing was not necessary for this age group.
Have you ever considered using out-of-level tests? Out-of-level tests will allow you to assess content mastery over time without encountering ceiling effects (i.e., students scoring at or near the 99th percentile on the pretest). You will learn what students do not know. You can document the challenge level of curriculum in your classroom, school, or district. If you currently use a pull-out program for several hours a week, you can also determine the extent to which time away from the regular education classroom affects mastery of concepts or principles. To what extent are students maintaining and enhancing their advanced-level skills?
Educators, parents, researchers, students, and the community at large want to be informed about students’ progress in the local schools. How are our students doing? Test data should serve various audiences. Resulting data aids decision-making about curriculum, instruction, and educational resources. Of course, test data over time is just part of the overall picture of how content, skills, and pedagogy come together in the learning process. Understanding the level of students’ daily performance is critical to planning and maintaining a strong focus on curriculum.
Educators, policy makers, and parents view tests as accountability measures. “Tests of student achievement that can be widely and uniformly administered across schools are the key mechanism by which policy makers hold schools accountable” (Donovan, Bransford, & Pellegrino, 1999, p. 54). Tests often serve as the barometer of local education achievements. School, district, and state reports provide considerable data about progress towards content standards or the percentage of students achieving at high, average, or low levels. Data may be portrayed over several years to show trend lines. At a glance, such portrayals provide information about preset goals. We have considerable experience in measuring factual knowledge and using objective scoring. We often make comparisons of the individual to a larger group of test takers of a similar age or grade.
As accountability measures, achievement tests must be selected based on their connection to the curriculum. To what extent does the scope and sequence outlined in your textbooks reflect the skills assessed on your school, district, or state level tests? Do the objectives of your curriculum reflect content standards in language arts, reading, science, mathematics, history, geography, or the arts? Are you measuring what is actually taught?
Given the availability of content standards developed by various professional organizations, it is easy to review the connections between curriculum and assessment. Note that it is curriculum and assessment, not curriculum then assessment. These processes are inextricably linked. As Elmore and Rothman (1999) state “the key is transparency” (p. 3). Administrators, teachers, students, parents, policy makers, and the community-at-large must know what is expected as outcomes of education, how outcomes will be measured, and how results will provide guidance about future learning opportunities. We must
The limited definition of tests offered above is not the only source of knowledge gained about student progress and instructional techniques. Our understanding of how people learn and how they transfer their learning is still unfolding. We are also very interested in deep understanding rather than surface, factual knowledge (Donovan, Bransford, & Pellegrino, 1999). We are gaining more and more expertise in measuring the depth of understanding. Newmann and Associates (cited in Elmore & Rothman, 1999) propose an emphasis on authentic pedagogy. They delineate four standards:
Deep Knowledge. Instruction addresses central idea of a topic or discipline with enough thoroughness to explain connections and relations and to produce relatively complex understandings.
Substantive Conversation. Students engage in extended conversational exchanges with the teacher or their peers about subject matter in a way that builds an improved and shared understanding of ideas or topics.
Connections to the World Beyond the Classroom. Students make connections between substantive knowledge and either public problems or personal experiences. (Elmore & Rothman, 1999, p. 75)
These four standards seem to be a good blueprint for thinking about the curriculum and assessment connections. They reflect and integrate viewpoints about testing:
Testing for Growth Over Time
Testing for Accountability
Testing and Performance
You’re having a test! The next time you say or hear this statement, ask yourself some critical questions about the purpose of the test, the scope of the questions, and how you will use the resulting data to improve the curriculum, change instructional techniques, or examine the strengths and abilities of your students.