Return to Index Page
Return to Chapter One

Chapter 4: Reading Assessment Profiles
Definition
Rationale
Reading Assessment: Principles and Trends

Chapter 4

Reading Assessment Profiles

Definition

Reading assessment is used to gather data to understand students' strengths and weaknesses in reading (Harris & Hodges, 1995, p. 12). This data is used to help design effective programs of instruction and to document the outcomes of instruction. The Tests of Adult Basic Education (TABE) and the Adult Basic Learning Examination (ABLE) are examples of two widely used, standardized tests in adult basic education that can provide teachers with information about at least two aspects of their students' reading: reading comprehension and vocabulary (CTB/McGraw-Hill, 1987, 1994; Karlsen & Gardner, 1986). Teacher observations and Informal Reading Inventories are examples of less formal measures. Teachers, test-makers, and researchers have all developed ways to assess students' reading ability in each individual aspect of reading: alphabetics, fluency, vocabulary, and comprehension. The characteristics of these methods are described separately in the major sections of this report.

Assessment profiles combine information from tests of several components to create profiles of learners' strengths and needs in reading for instructional purposes (Chall, 1994; Chall & Curtis, 1990; Roswell & Chall, 1994; Strucker, 1997). Learner profiling involves assessing a student in each major component of reading and using a common measure, such as grade equivalents, to create an outline of strengths and needs. When this technique is used, it is typically one of the first tasks a teacher completes and so it has been placed at the beginning of this report.



Rationale

Adult educators have traditionally used reading assessment to measure student growth in reading achievement and to diagnose individual strengths and weaknesses in reading in order to plan for instruction (Askov, Van Horn, & Carman, 1997). Determining what an individual learner or classroom of learners already knows and what they need to learn is thought to make instruction more efficient and effective. Measuring growth helps to determine whether a program of instruction has been effective (Askov et al., 1997; Joint Committee on Standards for Educational and Psychological Testing of the American Educational Research Association, American Psychological Association, and National Council on Measurement in Education, 1999).

Assessment profiles result in a comprehensive view of learner strengths and needs across all aspects of the reading process and can be used to design a program of instruction that addresses all aspects of the reading process during instruction. This ensures a balanced approach to instruction in which no one aspect of the reading process is over- or under-emphasized (NRP, 2000b; Snow, Burns, & Griffin, 1998; Snow and Strucker, 2000). In addition, researchers can use multiple measures of components to increase confidence that each assessment instrument is actually measuring the construct or aspect of reading that it is supposed to measure.



Reading Assessment: Principles and Trends


Question

Does assessment of adult learners' strengths and needs in reading instruction lead to increased reading achievement?

No trends are drawn from the research because there is very little experimental or non-experimental research that evaluates the effects of assessment on reading achievement, even at the K-12 level (Dochy, 2000 Segars & Buehl, 1999). It is widely assumed, however, that assessment of learner strengths and needs is an important aspect of instruction. In order to effectively and efficiently teach reading, a teacher must accurately assess an adult learner's ability in one or more areas of reading instruction (alphabetics, fluency, vocabulary, and comprehension). Reading assessment may be used to diagnose specific strengths and needs in reading for individual adults or for adults being taught in groups. It is also used to evaluate and modify instruction, and to evaluate overall ABE program effects on reading achievement.

Assessment Profiles


Question

Do assessments that include more than one aspect of the reading process, such as profiles, provide useful additional information for reading instruction? Based upon assessment profiles, what are ABE learners' strengths and needs in reading?

Several studies that describe the strengths and needs of ABE adult readers use more than one measure of reading ability (e.g., Greenberg, Ehri, & Perin, 1997; Scarborough, 1984; Pennington, Orden, Smith, & Green, 1990). Separate measures may be used for each important aspect of instruction: alphabetics, fluency, vocabulary, and comprehension. In some cases, more than one measure is used for the same aspect of instruction. Three measures for reading comprehension are used on the NALS, for example, each related to different content: prose, document, and quantitative (Kirsch, Jungeblut, Jenkins, & Kolstad, 1993). The studies described in this section focus on the profiles or patterns found across test scores in each area of instruction.

Principle 1:

When measures of achievement are obtained for each crucial aspect of reading instruction (alphabetics, fluency, vocabulary, and comprehension), instructionally relevant patterns of scores, or profiles of adults' strengths and needs in reading, may be observed. These profiles suggest that ABE readers, including those in ESOL programs and those with a reading disability, are very diverse and that any one measure of reading achievement may not be sufficient to identify strengths and needs for instruction. (Carver & Clark, 1998; Chall, 1994; Norman & Malicky, 1987; Norman, Malicky, & Fagan, 1988; Strucker, 1995, 1997).

In one non-experimental, descriptive study, a group of approximately 100 ABE learners were administered tests of word analysis and word recognition (alphabetics), oral reading (fluency), spelling, vocabulary, and comprehension (Chall, 1994). The learner profiles, or patterns of grade equivalent scores across the six measures, were analyzed. Two common patterns were found. One pattern describes ESOL adults, and the other seems to be similar to the patterns of scores found among children with a reading disability. For the ESL group, alphabetics and fluency scores are relatively high while vocabulary and comprehension scores are relatively low. For the reading disabled group, a different common profile was found. Print-based aspects of reading (alphabetics and fluency) tend to be relatively low while meaning-based aspects of reading (vocabulary and sometimes comprehension) tend to be relatively high. These two profiles were observed across ability levels, from beginning to advanced levels of reading.

Two studies of ABE learners' reading achievement profiles used factor or cluster analyses to look for common patterns across learners. In one (Norman et al., 1988, with a non-statistical presentation in Norman & Malicky, 1987), the following scores were used to generate profiles for over 100 adults reading at Grade Equivalent 1 through 8 (GE 1-8): reading comprehension and word recognition achievement scores on an informal reading inventory, the number of oral reading miscues in five miscue categories, and the number of ideas (clauses) recalled in four separate categories. Miscue and clause categories were based on the degree to which miscues and clauses were either text-based or knowledge-based. An oral reading miscue that resembles a word in the original text (saying "bark" when the word in the text is "dark," for example) is text-based, while a miscue that does not is presumed to come from the reader's knowledge base. Similarly, a recalled idea that closely resembles one in the text is text-based, while one that does not is knowledge-based. Results from the analysis suggest that profiles cluster into two or three groups, based on developmental stages in reading ability. Beginning readers (GE 1-4) attend more to the print on a page than they do to their own knowledge as they read. More advanced readers (GE 5-7) rely on both print and knowledge and are better able to integrate the two. An intermediate or transition group at about GE 4 may also be present.

Another cluster analysis study is described (in Strucker, 1997) along with two case studies of adult learners' reading profiles (quantitative details of the cluster analysis are reported in a non-peer-reviewed manuscript, Strucker, 1995). Data from seven measures of reading for over 100 ABE learners were used in the cluster analysis. These included measures of alphabetics (phonemic awareness, word analysis, and word recognition), spelling, fluency (oral reading), oral vocabulary, and reading comprehension. In general, two categories of profiles were identified, ESL and reading disabled, supporting observations made in one of the descriptive studies cited above (Chall, 1994). In addition, nine patterns of strengths and weaknesses in reading that may be instructionally relevant were found across several developmental levels or stages. There were two profiles for beginning readers (GE 0-3): beginners and advanced beginners. There were four profiles at the intermediate level (GE 4-8): ESL and non-ESL inner city young adults, ESL and non-ESL reading disabled, reading disabled (low comprehension), and ESL and inner-city young adults. The three remaining profiles were at the advanced ABE level (GE 9-12): reading disabled (high comprehension), pre-GED low vocabulary, and GED high vocabulary.

A series of three studies allows a comparison of the reading profiles of ABE students with those of children and advanced adult readers and supports the use of profiles to describe adults' strengths and needs in reading (Carver & Clark, 1998). Children finishing the 3rd, 4th, and 5th grade, a group of normally achieving university freshman, and a group of 128 average to poor readers from a community college were all assessed with the same computer-assisted assessment system that measures oral and silent vocabulary, reading comprehension, oral reading rate, word recognition, and the speed at which readers can name letters. Reading profiles derived from these scores suggest that the average scores of children and university freshman form flat profiles, with roughly the same GE scores for all components, while poorer adult readers have noticeable strengths and weaknesses (ups and downs) in their profiles. Averages scores for the various components assessed were all in the GE 12 to GE 13 range for the university students and at GE 5 for the children. GE scores in the profiles of the poor readers from the community college were, on the other hand, variable, with rate (fluency) and word recognition (word analysis) scores usually being the lowest scores. A more qualitative description of the subset of the community college group with a reading disability, defined as having at least one GE score of 6 or lower, suggests that most (98%) of these reading disabled adults had low rate scores, or a rate disability. All also had low word recognition scores, and 67% scored below GE 6 on a measure of oral vocabulary knowledge.

Back to top
Back to Chapter One
Next