Abstract
During the past 5 years, a variety of assessment procedures have been labeled “curriculum-based assessment” (e.g., Blankenship, 1985; Gickling, 1981; Shapiro & Lentz, 1985). Among these is a methodology referred to as Curriculum-Based Measurement (CBM) (Deno, 1985, 1986; Deno & Fuchs, 1987). CBM qualifies as a form of curriculum-based assessment because it meets the following three requirements outlined by Tucker (1987): (a) Measurement materials are drawn from individual students’ curricula; (b) measurement is ongoing; and (c) assessment information is used to formulate instructional decisions.
Beyond these features, which are common across various forms of curriculum-based assessment, CBM is distinctive because of two additional properties. First, within the framework of an academic year, CBM is conducted on samples from an unchanging pool of stimulus materials; this pool comprises the year-long curriculum. Second, CBM is highly prescriptive and standardized.
For example, with most forms of curriculum-based assessment, teachers measure students’ mastery of a sequence of short-term instructional objectives. The practitioner specifies the sequence of skills to be mastered, applies broad guidelines to develop his or her own criterion-referenced measurement procedure to match each skill, and initiates instruction and measurement on the skill presumed to fall lowest on the instructional hierarchy. When students attain mastery of the first skill, the teacher simultaneously shifts instruction and measurement to the next objective in the sequence.
By contrast, practitioners implement CBM by specifying the pool of material representing the entire year’s curriculum. This pool constitutes the domain of the long-term goal; that is, by year’s end, the student will demonstrate proficiency on the entire pool of material. Teachers use standardized procedures for sampling test stimuli from that pool and for administering, scoring, and using these tests to assess student proficiency on the year’s curriculum.
Although CBM has been described in the special education literature, there is relatively little explanation of how and why CBM differs from other forms of curriculum-based assessment. Given that research demonstrates that (a) CBM is a reliable and valid form of measurement for a variety of assessment-related decisions (Deno, 1985; Germann & Tindal, 1985; Marston, 1988; Shinn, 1989) and that (b) teachers can use CBM to assist in effective instructional planning (Fuchs, Fuchs, & Stecker, 1989) and in enhancing achievement for learning disabled and other academically deficient pupils (Fuchs, Deno, & Mirkin, 1984; Fuchs, Fuchs, & Hamlett, 1989a), a comprehensive description of how CBM can be used to help teachers formulate effective instructional programs seems appropriate at this time. Consequently, the purposes of this article are to explain the rationale for CBM’s standardized, long-term goal approach to measurement and to describe how CBM can be used to develop instructional programs.
Rationale for a Standardized, Long-Term Goal Approach to Measurement
As an approach to curriculum-based assessment, CBM’s two distinctive features are (a) standardized measurement and (b) assessment of proficiency on long-term goals. A rationale for each dimension follows.
Standardization
When curriculum-based assessment requires teachers to design their own assessment procedures to match a series of short-term objectives from an instructional hierarchy, two disadvantages exist. First, curriculum-based assessment is more time-consuming when the design of measurement procedures (a) changes each time a student masters an objective and (b) differs across pupils in the same classroom. Second, the adequacy of teacher-developed tests is unknown; that is, we are uncertain of the degree to which scores on teacher-made tests represent performance on meaningful, important skills (i.e., validity), and the extent to which the student would achieve a similar score if the test were readministered (i.e., reliability).
Moreover, even when teachers use commercial criterion-referenced tests for their curriculum-based assessment, the meaningfulness and accuracy of scores frequently is questionable: Among 12 commercial criterion-referenced instruments, authors of only 4 addressed reliability or validity, and usually with only one index of adequacy (Tindal et al., 1985). Additionally, independent analyses of criterion-referenced tests attached to basal reading series reveal varying degrees of reliability and validity, with many estimates falling considerably below acceptable levels (Tindal et al., 1985).
By contrast, CBM relies on an alternative strategy for the design of measurement procedures. We have investigated a variety of measurement procedures to determine a set of steps that can be applied to any typical reading, spelling, math, or written expression curriculum (see Shinn, 1989, for summary of research). Regardless of the curriculum employed, these standardized procedures produce meaningful and accurate scores. (The specific measurement procedures have been outlined previously; see, for example, Deno & Fuchs, 1987; Fuchs, 1987; Shinn, 1989. They are not repeated here.) The advantages of employing these prescriptive, standardized procedures are threefold: The time-consuming burden of developing measurement procedures is lifted from the teacher, the process for measuring student performance within an academic area remains constant across time for a given student and across different pupils, and the teacher can be confident in the meaningfulness and accuracy of the scores.
Long-Term Goal
Within the context of CBM, two assumptions about IEP goals are critical. First, we assume that IEP goals should describe important annual outcomes that incorporate indicators of overall student proficiency. Stated another way, IEP goals should not describe numerous subskills or provide overviews of instruction to be provided. For example, over the course of an academic year, we might plan to focus beginning reading instruction on a series of skills (e.g., sound-letter correspondence; reading consonant–vowel–consonant [CVC], CVCe, CVVC, and r-controlled words; identification of characters, setting, main events, and sequence within narrative text; sight vocabulary; fluency). Teaching these skills may represent key instructional events in effecting student progress. However, we do not specify IEP goals (and corresponding IEP-monitoring strategies) so that one goal addresses each skill. Rather, we state one IEP goal for the entire year, in terms of one student behavior that simultaneously represents all the skills we want the student to master.
For example, given a beginning reader, our CBM research (e.g., Deno, Mirkin, & Chiang, 1982; Fuchs, Fuchs, & Maxwell, 1988) indicates that a valid student behavior, which simultaneously reflects numerous reading skills, is oral reading from the Grade 1 basal text. Proficient reading in this text requires decoding, sight vocabulary, fluency, and comprehension: When students lack skills, scores are low; as students master these skills, scores increase. Consequently, we do not need to list every subskill in our goal, because our curriculum-based test of overall proficiency (i.e., oral reading from text) indexes the extent to which the student has mastered the skills embedded in the first-grade curriculum. Our basic assumption is that, when the IEP goal incorporates global annual outcomes, IEPs become more manageable and useful documents for monitoring student progress. In fact, given this first assumption, the IEP current level of performance, goal, and objective within an academic domain can be stated as three parallel sentences. Table 1 shows sample IEP statements in three academic areas.
CBM Current Performance Levels, Goals, and Objectives for Reading, Spelling, and Math.
A second key CBM assumption is that the primary purpose of the IEP goal is to structure the monitoring of student progress. When a global annual outcome is specified as the IEP goal, each monitoring test incorporates the entire set of subskills embedded within the curriculum. So, given the IEP reading goal in Table 1, CBM requires the student to read passages on which we expect proficiency at the end of the year. In October, when the student has mastered only a subset of the skills necessary for proficient Grade 1 reading, the CBM score is a relatively low 16. Figure 1 represents the mastered skills (shown in small boxes) that might be associated with a low CBM score of 16 (see score at bottom of Figure 1). But as the student masters more of the curriculum, the associated CBM score increases (see Figure 2, which represents the mastered skills [shown in small boxes] that might be associated with a higher score of 71, earned by this student later in the year).
Mastery of skills (highlighted in the box) that are reflected in a relatively low CBM oral reading fluency score on first-grade material.
Mastery of skills (highlighted in the box) that are reflected in a higher CBM oral reading fluency score on first-grade material.
The information embedded in the current level of performance, goal, and objective is transformed easily into a graphic display (see math example in Figure 3) to structure the monitoring of student progress: The first dots represent the current level of performance, the G indicates the goal, and the broken line signifies the objective (representing the rate of expected progress and called the “goal line”). Consequently, the IEP goal provides the structure for the curriculum-based progress-monitoring system. Each score is an overall indicator of student proficiency in the curriculum. As scores increase, we know progress is occurring; if scores stay the same, progress is lacking. The result is a manageable IEP document that provides a relatively efficient structure for monitoring overall student progress.
A CBM math graph showing a performance pattern indicating that an increase in the goal is necessary.
How CBM Can Be Used to Develop Instruction
CBM databases can be used to develop instructional programs in four ways: (a) to judge the appropriateness of the goal, (b) to assess the adequacy of student progress and the need to modify an existing program, (c) to contrast the efficacy of different treatments, and (d) to determine potentially effective strategies for modifying and enhancing instructional programs.
Judging the Appropriateness of the Goal
CBM databases can be used to judge the appropriateness of the goal in the following way: When student progress indicates that the student’s actual rate of progress exceeds the expected rate, teachers increase the goal. For example, given the graph shown in Figure 3, the student’s initial performance was 25 digits in the fourth-grade curriculum, and the teacher set a math goal of 45 digits correct. The teacher sets the goal, institutes the first instructional program (indicated by the broken vertical line), and measurement continues, with each test representing all the problem types of the fourth-grade curriculum. As shown in Figure 4, the student’s actual rate of progress (solid diagonal) is greater than what was anticipated (as shown in the broken goal line). Consequently, the teacher raises the goal, draws a new broken vertical line to symbolize the goal change, draws a new goal and goal line, and continues to measure.
A CBM math graph, showing a performance pattern that indicates a teaching change is required.
This CBM process is facilitated by software (Fuchs, Hamlett, & Fuchs, 1990) that automatically generates, administers, and scores tests, saves student scores and responses to the items on the tests, graphs scores, and automatically analyzes the student’s rate of progress in comparison to the goal line. To raise the goal, the teacher simply indicates to the computer a new goal of 60, and the computer automatically draws the broken vertical line, a new goal (represented by G), and a new goal line (signified by a broken diagonal). See Figure 4.
In an investigation of the effects of using this “dynamic” goal-setting process in the area of math (Fuchs et al., 1989a), half of the participating CBM teachers modified goals as just described, in response to students’ actual progress. The other half were free to change goals whenever they deemed it appropriate; however, the computer never directed them to increase goals. Fuchs et al. found that teachers who employed the goal-increasing decision rule increased goals for 60% of participating students, whereas nondynamic goal-setting teachers raised goals for only 1 of 20 students. Results indicated that teachers frequently set initial goals too low, but that, without decision rules to increase goals, teachers tended to maintain inappropriately low goals. This pattern of low goal setting has been corroborated in additional research in reading, spelling, and math (Fuchs, Fuchs, & Hamlett, in press; Fuchs, Fuchs, Hamlett, & Allinder, 1991; Fuchs, Fuchs, Hamlett, & Stecker, 1990). Fuchs et al. (1989a) also found that, with dynamic goals and resultingly higher levels of goal ambitiousness, teachers effected better student achievement. Consequently, using CBM databases to develop goals dynamically may enhance teacher decision making and student achievement.
Assessing the Need to Modify Programs
A second way in which CBM can assist teachers in their instructional program development is to judge the adequacy of student progress and the need to change instructional programs. The following decision rule, applied to CBM databases, structures this process: When (a) the line of best fit drawn through at least eight scores is less steep than the goal line (as shown in Figure 4) or (b) at least four consecutive scores fall below the goal line, the student’s progress is judged inadequate and the computer recommends a teaching change.
Research indicates that this type of instrumental use of CBM databases is important to student achievement. Fuchs, Fuchs, and Hamlett (1989b) assessed reading achievement using commercial standardized reading comprehension tests and using the slope of the CBM database for two groups: (a) 21 students whose teachers used the CBM database instrumentally to formulate instructional changes and (b) 15 students whose teachers measured student performance but failed to use CBM data to determine when programmatic improvements were necessary. On both achievement indices, students in the first group achieved better. This finding corroborates previous work (Fuchs & Fuchs, 1986) suggesting the need for systematic use of CBM data to enhance student achievement outcomes.
Contrasting the Efficacy of Different Treatments
CBM also can be used to compare the efficacy of different instructional programs or service delivery options. As shown in Figure 5, the intervention implemented during November and December produced little growth compared to the baseline level of performance. In January the teacher introduced a second intervention, which resulted in dramatically better growth. In March the teacher began to experiment with a third intervention, which initially produced even better progress compared to previous teaching programs. Based on the CBM data, the teacher can determine which relatively effective instructional components to incorporate in subsequent interventions and which less effective dimensions to eliminate from consideration. In this way, teachers can experiment with alternative instructional programs for individual students and judge their efficacy not only in relation to an arbitrary standard established in the goal, but also with respect to competing interventions. Deno (1985) refers to this process of experimentation with different program components as “experimental teaching” (see Casey, Deno, Marston, & Skiba, 1988). Marston (1988) provided an interesting application of experimental teaching, in which the efficacy of regular and special education service delivery options were compared.
A CBM math graph, showing a student’s rate of progress under three different interventions.
Determining Strategies to Modify and Enhance Instructional Programs
Within the first three strategies for using CBM to improve instruction—judging the appropriateness of goals, assessing the adequacy of student progress along with the need to introduce instructional improvements, and comparing the efficacy of different interventions—CBM practitioners rely solely on the graphed database, which incorporates the total scores earned on the CBM tests. These total scores represent overall indicators of student proficiency within the target curriculum, and are especially useful for decisions that involve judgments about overall improvement, such as decisions about when to raise goals, when to modify existing programs, and the efficacy of competing programs. Yet, additional information is available from the CBM database to help teachers determine how to modify and enhance students’ programs. For this last purpose, practitioners can rely on an analysis of the students’ responses to the actual CBM test items.
When analysis of the overall performance indicators reveals that the student has not been making adequate progress, the teacher introduces a teaching change. To get additional description about the student’s performance in the curriculum, the teacher can look at the computerized skills analysis. As shown in the top part of Figure 6, the computer groups the objectives within the fourth-grade curriculum into those the student currently has Mastered, Partially Mastered, Nonmastered, and Not Attempted. In the Performance History, shown in the bottom part of Figure 6, boxes are pattern-coded to indicate the student’s category of mastery for each objective, for each 2-week interval across the school year. Black boxes indicate Mastery; checked, Partial Mastery; striped, Nonmastered; and white, Not Attempted. Consequently, the teacher looks for increasingly darkened boxes across time to indicate improvement.
A CBM skills analysis in the fourth-grade curriculum.
Research on teachers’ use of skills analysis within CBM is promising. Studies were conducted in reading, spelling, and math. Half of the CBM teachers in each academic area used only the graphed analysis of the performance indicators; the other half used the graphed analysis along with the skills analysis. In all three academic areas, addition of the skills analysis resulted in more specific instructional program descriptions and better achievement for learning disabled and other low-achieving pupils (Fuchs et al., 1991; Fuchs, Fuchs, et al., 1990; Fuchs, Fuchs, & Hamlett, 1989c). Consequently, it appears that all four strategies for using CBM data to improve instructional programs make important contributions to the achievement of handicapped learners.
Summary
CBM is a form of curriculum-based assessment that employs a long-term goal measurement strategy and a prescriptive, standardized measurement methodology. This focus on global annual outcomes within the curriculum, using measurement strategies with documented reliability and validity, results in more efficient, accurate data-collection activities. Additionally, research demonstrates that the resulting database can be used by special educators to enhance the quality of instructional programs delivered to learning disabled and other low-achieving youngsters. This improvement in instructional planning and student achievement can be accomplished in four ways: (a) by judging the appropriateness of the goal and adjusting goal levels in response to the CBM database, (b) by monitoring the adequacy of student progress as indexed by the graphed CBM data and adjusting instructional programs when student performance patterns indicate inadequate improvement, (c) by comparing the efficacy of alternative programs using the CBM graphed database and building on effective components while eliminating ineffective program dimensions, and (d) by using analyses of student responses to the CBM test items to determine strategies for improving instructional programs.
Footnotes
Editors’ Note
The third decade of Intervention in School and Clinic, then titled Academic Therapy, represented a decade focused on research and its translation into classroom practice. The field began its mission to ensure that every student received an education based on research. We read articles in the archives by Dr. Susan Goldman, Dr. James Pellegrino, Dr. James Patton, Dr. Maggie Coleman, Dr. Jo Weber, Dr. Ann Ryan, Dr. Ed Ellis, and Dr. Ginger Blalock. All of these researchers were working diligently to translate their research into practical classroom strategies. We selected an article by Dr. Lynn Fuchs, Dr. Douglas Fuchs, and Dr. Carol Hamlett, all at Peabody College, Vanderbilt University, at the time of publication. The article provides an important overview of curriculum-based measurement and its implementation in the classroom. The suggestions and strategies are as timely today as they were in 1990.
Authors’ Note
Statements should not be interpreted as official agency positions.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research described in this article was supported, in part, by Grants G008530198 and G008730087 from the U.S. Department of Education, Office of Special Education Programs, to Vanderbilt University.
This paper was originally published as: Fuchs, L. S., Fuchs, D., & Hamlett, C. L. (1990). Curriculum-based measurement: A standardized, long-term goal approach to monitoring student progress. Academic Therapy, 25(5), 616-632.