Since 1977, significant advances in anthropology, cognitive psychology, education, linguistics, and sociology have made it possible to expand how reading is viewed. These advances indicate that reading is a dynamic process in which the reader actively participates. As a result, difficulty is no longer viewed as a property of a particular reading skill or task, but rather as an interaction among the reader, text, and context of the reading situation.

This article looks at authentic reading assessment as a response to this evolving concept of reading. An examination of the Michigan State Board of Education's reading assessment--one of the nation's most innovative--shows how one state is implementing authentic reading assessment using authentic reading material.

MICHIGAN'S ESSENTIAL SKILLS READING TEST

In 1983, the Michigan Department of Education adopted this definition of reading:

"Reading is the process of constructing meaning through the dynamic interaction among the reader's existing knowledge, the information suggested by the written language, and the context of the reading situation."

This definition suggests the need for an interactive model of reading which combines the top-down (whole language) and bottom-up (skills) models. This interactive model leads to the skills taught in the context of real reading of real text.

From this perspective, a good reader is one who can apply various reading skills independently and flexibly in a variety of reading situations, not one who simply demonstrates mastery of those skills.

Given this theoretical focus, a variety of factors needs to be addressed in instruction and assessment. These include:

* the influence of the reader's prior knowledge on reading comprehension,

* how the reader structures that knowledge,

* which strategies the reader uses to construct meaning,

* which skills the reader needs to perform a particular reading task,

* the type of methods and materials being used, and

* the setting in which reading occurs.

These are not new concerns, but reading research and theory have only recently enabled educators to integrate these issues into instructional and assessment practices.

Michigan's "goal" for reading education is to develop strategic motivated readers. This goal highlights the importance of a reader having knowledge of the reading process, knowledge about strategies and skills that are essential to constructing meaning or comprehending text, and knowledge about how to appropriately apply these strategies. The goal also includes motivation to read.

Unlike other formal assessments of reading, Michigan's Essential Skills Reading Test uses intact, full-length stories and subject-area reading selections taken from real life materials, such as children's magazines, literature anthologies, and textbooks for different grade levels. The reading selections are then the driving force for developing test items.

TYPES OF ITEMS ON THE TEST

The Essential Skills Reading Test has three types of constructing-meaning items:

1. Intersentence, in which the answer to this type of item can be found in two to three contiguous sentences within the reading selection;

2. Text, where one or more paragraphs of the reading selection must be read to construct meaning; and

3. Beyond Text, where the reader not only constructs meaning from the text but also must bring in some of his or her own prior knowledge to answer the test item.

In addition to the constructing-meaning items, knowledge-about-reading items might ask students if they know the purpose of the chart, graph, or illustration. Other items of this type might ask students whether they know how a reading selection is organized and how that organization influences comprehension. This can include items about what strategies should be considered when readers encounter various unknown situations.

Student-self-report items specifically ask the students about their interest in the reading selection, how they felt about their performance in understanding the reading selections and answering the test items, and the amount of effort they put into reading the selections and answering the test items. Like the knowledge-about-reading items, these items also are specific to the reading selections in the test.

REACTIONS OF PARENTS AND TEACHERS

Overall, reactions to the Michigan test have been positive. Many welcomed the long due change in the approach to assessing reading performance, although some were not as eager to adopt the new approach. Change is occurring, but perhaps not as rapidly as had been hoped.

Michigan officials made extensive efforts to inform the public about the new tests and their purpose. They prepared people for the results by emphasizing that "low scores are good news." As a result, when the low scores arrived, many parents and teachers accepted them as an opportunity to improve education in Michigan.

CURRENT RESEARCH IN READING ASSESSMENT

The impact of the continuing research on reading and reading assessment is reflected in the 1992 National Assessment of Educational Progress (NAEP) reading objectives framework. This framework introduces several new innovations to reading assessment, such as:

* developing a large proportion of open-ended items that assess constructing-meaning, including essay questions;

* expanding the constructing-meaning scale to include three different levels (building an initial understanding, developing an interpretation, responding personally or critically);

* including reading for three purposes: to gain literary experience, to get information, and to perform a task; and

* providing a measure of reading fluency or oral reading proficiency.

The new framework will continue the past emphasis on assessing the use of effective strategies, increasing knowledge about reading, and developing positive reading habits and practices. It will use longer, naturally occurring passages. The results will be reported on three scales, one for each reading purpose cited above, and will be summarized on one overall scale.

In Michigan, the Reading Test Development Coordinating Committee is in the process of reviewing and revising its Essential Skills Reading Test. Some of the issues the committee is considering include:

* formats other than multiple-choice (such as open-ended items) and other types of reading materials (such as those related to employability skills)

* test length--whether the current selections are too long and what length would be sufficient

* asking questions across tests to see if students can use more than one source of material to gather information and make decisions

* classroom assessment

ADDITIONAL READING

Dutcher, P. (1989). Reading the MEAP Test Report Forms.

Dutcher, P. & Roeber, E. (1989). New MEAP Reading Test Debuts. Michigan Association of School Boards, 50, 9.

Michigan State Board of Education (1988). Essential Goals and Objectives for Reading Education.

Michigan State Board of Education (1989). Essential Skills Reading Test Blueprint, (Fifth Edition).

Roeber, E. & Dutcher, P. (1989). Michigan's Innovative Assessment of Reading. Educational Leadership, 46, 7.

Wixson, K. & Peters, C. Reading Redefined: A Michigan Reading Association Position Paper.