THE Learning Record ASSESSMENT SYSTEM

by Mary A. Barr
Center for Language in Learning

This paper describes the Learning Record (LR) system, which

The CLR system has undergone six years of research and development throughout California, the last two at the Center for Language in Learning, a not-for-profit organization in El Cajon, California. From 1991-1994, the work was carried out under the aegis of the Chapter 1 office at the California Department of Education through the University of California at San Diego. The CLR has been adapted, with permission, from the Primary Language Record, which originated in inner London in the early 1980's, extending it to all subject areas and grade levels. By June, 1994, all parts of the system had been tested in the U.S. in small scale studies so that it could begin to be phased in by Chapter 1 program schools as an alternative or complement to norm-referenced, multiple choice testing. The Center continues to monitor school and district results of CLR use and its quality assurance program.

The CLR system of assessment is standards-referenced. Unlike norm-referenced and criterion-based assessments which rely on single scores obtained from performance on standardized tasks, standards-referenced assessment requires analysis of a pattern of performance observed and documented over time in natural settings. Descriptive statements about the degree of achievement are drawn from agreed upon performance standards and scales, multiple perspectives with documented observations of how those standards are being met, and exemplars which illustrate a variety of ways teachers have provided evidence that individual students have met the standards.

Useful for both improvement of teaching and learning and for summative reporting, the CLR is a classroom-based student assessment system which combines the use of external public descriptions of levels of student performance with individual definitions of successful achievement to produce an annual record of achievement. Interrelated and overlapping safeguards ensure equity and reliability. This kind of assessment moves away from post hoc review and judgment of student work in favor of the direct classroom use of performance standards in regard to student work in progress as well as to the products of that work. Judgment is framed in terms of criteria and standards which can be used jointly by teacher, student and (especially with elementary aged students) parent.

Sources of Data

The complexities of learning situations have long been oversimplified in order to accommodate the requirements of standardized assessment tasks and, consequently, rich sources of information about the student's learning have been lost. In contrast to conventional assessment systems which seem to encourage only low level skills, the CLR system of assessment encourages the use of natural settings to support learners' increasing abilities to solve problems in multiple ways, to interpret text, to communicate their interpretations and to express informed opinions.

With the CLR, the focus is on the collecting and documenting of abundant evidence of what students know and can do in natural settings. Teachers collect and organize information about students' learning from multiple sources, using the principle that student use of language and literacy is fundamental to academic progress across the curriculum. They collect and document evidence from which to make judgments about student achievement throughout the year; students and parents contribute to the accumulating data for making judgments. The assessment is integrated with classroom activities and projects so the information is available to serve teaching and learning. In addition, it is possible to report the numerical achievement results directly from the school to the broader community for aggregating and summarizing of student achievement and school program results.

The schematic below depicts this systematic flow from classroom to public accountability in the CLR assessment system. It illustrates how the perspectives of parent, student and teacher contribute to a multi-faceted picture of achievement. A standardized format is used so the picture can be interpreted across school, district and state levels. The picture is both qualitative and quantitative, with placement on performance scales determined by the documented evidence obtained throughout the year from teacher observation as well as from samples of student work.

schematic diagram of CLR system

Assessment tasks are the same as learning tasks, chosen or designed by the teacher as appropriate for his/her students. Consequently, teachers do not stop teaching and students do not stop learning in order to prepare for testing. Standardized tasks prepared externally to be administered under test conditions may also be used, but only at the discretion of the teacher. Assessment is dynamic, informed by an on-going program of professional dialogue at the site or in the region, by the application of familiar and open performance criteria, and by the aforementioned exemplars of student work from schools throughout the state at different grade and proficiency levels. Teachers participate in regional readings of records in order to validate their judgments as accurate and fair. At these readings they compare and mediate their interpretations of the collected evidence in each student's record.

Procedures for Assuring Equity, Validity, and Reliability

The CLR approach to assessment recognizes the need for both qualitative and quantitative judgments about student performance. In order to ensure these judgments are fair and accurate in classrooms as well as in cross district settings, the CLR system supplies interrelated safeguards within the Learning Record itself, in the implementation guidelines for teachers and administrators at each school site, and in the program for public accountability in each region:

Specifically, the system uses numerical as well as non-numerical descriptions of five levels of student performance in reading, writing and mathematics at Grades K-3, 4-8, 9-12. The former permits teachers to quantify achievement data about performance with numerical cutoffs established at given grade levels. For example, the cutoff established in primary reading is at the end of Grade 3, that is, by that time all students should be at a 4 or 5 on the scale. The numbers on the performance scales represent gradations of achievement which, in combination with narrative descriptions of performance and examples of student work, provide both formative and summative information. For example, the primary level performance scale in reading describes students in stages as they become readers. This is summative information, which can be used to define the percentages of students at various stages of becoming readers at the school, regional and state levels. The accumulating evidence about individual student reading performance in Grades K-2 allows teachers and parents to monitor and support students as they move through a series of developmental benchmarks along the journey from being a dependent reader to an independent one. Used as formative information, the CLR identifies not only which learners need help but what they need help to do.

The evidence collected each year, to be considered valid and reliable, has to persuade other teachers who teach at the same grade levels to place the student similarly. In small trials in California, where we masked the placement information of the classroom teacher and asked other teachers using the CLR to read the student record together with samples of student work analyzed in the record, we have been able to reach a high level of interrater reliability. We attribute this consistency to the fact that two statewide school support networks collaborate to assure agreement among the state's teachers in regard to interpretations of performance criteria and to help make educational standards accessible not only to teachers but to students and parents.

Support Networks

The Center for Language in Learning Coordinates networks with teacher leaders in Core Development Groups who inform and support classroom-based assessment in the following ways:

Center staff

To sum up, the Center for Language in Learning seeks to make available for students and their parents a balanced, fair, and accurate alternative system of assessment which also serves district and state accountability purposes. The California Learning Record with its program of quality assurance promises to do just that.

The Center for Language in Learning, 10610 Quail Canyon Road, El Cajon, CA 92021, produces and monitors the Learning Record Assessment System, K-12.

Note: The term "standards-referenced assessment" is one borrowed from D. Royce Sadler, Oxford University, to describe an approach to assessment which makes "direct use of teachers' qualitative judgments" and which aims to link student grades and documented evidence of achievement as indicators of patterns of performance on multiple tasks over time. (See the full discussion in "Specifying and Promulgating Achievement Standards")

Questions? Email Dr. Mary Barr at lrecord@cll.org


© 1995-2006 M. A. Syverson