by Jay Howard
Developmental teachers need just the right words to describe the reading abilities of their students. They have to use terms that are appropriate to the task, reading in college, and which inform students, other teachers, and the public. The words should be precise enough to distinguish developmental college readers from others, but not so fine that they constrain instruction or are morphed into jargon. An examination of the words used will likely reveal an underlying educational philosophy, and the users of the terms should be aware of the connotations or implications that these words convey.
The terms used to characterize reading may vary between two illustrative extremes. Some terms refer to elementary grade levels, for example, "He is reading at the fifth-grade level." The public may recognize these terms, but it could be asked how they fit into the scope of reading skills at the college level or just what does "fifth-grade level" mean? Is that the fifth grade at Masterman across the street? Or is it a reading group in a fifth grade class at Waring, two blocks away from the main campus and with a different student population?
While the grade level qualifiers may appear to have precision, other terms may have a creative quality that makes them appealing. In a book on City College of New York, Traub (1994) quoted a dedicated developmental teacher who said that his students engaged in "...daydreaming-type thinking" (p.140) This oxymoronic expression sounds as if it is capturing the essence of a unique behavior, but just what does it mean? Some terms may be so evocative that they do not provide much useful information.
The educators who made up the National Adult Literacy Survey, or NALS, were faced with this problem. They had to find terms to describe reading abilities for adults who demonstrated a wide range of reading abilities. Since their results have been published, and since they surveyed over 1,000 community college graduates, it might be helpful to consider what terms they used to describe reading and to consider how the literacy levels these terms describe might distinguish developmental reading from "college" reading.
The NALS was developed jointly by the National Center for Educational Statistics, the Division of Adult Education and Literacy, and the Educational Testing Service. The test developers emphasized the functional aspects of literacy and "... endorsed the idea that literacy involves a multiple set of skills arrayed along a continuum, rather than discrete skills that are context bound and that literacy skills are employed across a variety of adult contexts that include work, home, and community" (Campbell, Kirsch, & Kolstad, 1992, p. 10). These criteria were not linked to the college classroom, but given the scope of the NALS, it might still be useful to consider how they defined reading.
The NALS divided text literacy into three general categories: prose, document, and quantitative. Prose and document literacy will be considered here. Prose materials were reprinted in toto from sources such as books, magazines, or newspapers. The document materials came from tables, charts, graphs, or maps. For each scale a series of open-ended questions was written. The questions were scored on a 500 point scale and divided into five levels, from the first basic level to the fifth.
The literacy tests were given by over 400 trained interviewers to over 26,000 individuals over the age of 16. About 4% of the total, or 1,033 participants, said that they had completed two years of college.
For both the prose and document text, at Level 1, the reader had to locate a discrete bit of information that was similar to the text of the question. The sample question asked the reader to underline a sentence from a short newspaper article or fill in personal information on a simple form. From the total population, 21% of the adults scored at this level; 4% of those with associate's degrees were at Level 1 on the prose and 6% on the document scales.
At Level 2, the participants were asked to complete the tasks for Level 1, but they might also have to form simple inferences or compare or contrast some information in prose materials. Instructions for operating equipment were given as a sample. On the document materials, they had to match, form inferences, or integrate data from different sections of a document. A pay check stub and a line graph showing sales over three years were included in the document test. Over a quarter (27%) of the adults scored at this level on the prose scale and 28% on the document scale. Among the community college graduates, 19% were at this level on the prose scale and 23% on the document scale.
Level 3 increased the demands. The sample text was a long newspaper article and the respondents had to do all the preceding as well as "integrate information from ... text that contains no organizational aids, such as headings." (Kirsch, Jungblut, Jenkins, & Kolstad, 1993, p. 78.) The largest group of the general population, 32%, and of the community college graduates, 41%, were at this level. "Tasks within the range for Level 3 [document] ask the reader to locate particular features in complex displays, such as tables that contain nested information" (Kirsch, Jungblut, Jenkins, & Kolstad, 1993, p. 89). Again, the largest group of the general population, 31%, as well as the two-year grads, 43%, were at this level.
For Level 4, the readers had to "... perform multiple-feature matches and to integrate or synthesize information from complex or lengthy passages. More complex inferences are needed to perform successfully... [and] conditional information is frequently present." (Kirsch, Jungblut, Jenkins, & Kolstad, 1993, p. 80) Two newspaper articles were given as examples. Only 17% of the adults but 32% of the community college grads were at this level. The document sample was a bus schedule and the reader had to integrate a good deal of information and make inferences. Of the adults, 15% were at this level, as were 25% of the associates degree holders.
The sample of the most difficult materials for Level 5 was taken from a handout given to prospective jurors. The readers were asked to search for information in text with plausible distracters, make complex inferences, or use detailed background information. Just 3% of the adults and 4% of the community college graduates were at this level.
For the document materials, the sample was a table from an opinion survey and the respondents had "... to search through complex displays that contain multiple distracters, to make high-level text-based inferences, and to use specialized knowledge" (Kirsch, Jungblut, Jenkins, & Kolstad, 1993, p. 92). Again, 3% of the adults were at this level, as were 3% of the community college grads.
The terms used in the NALS to describe each level of literacy seemed to be based on a general understanding of what constituted reading. In the field of reading, there are a number of information processing models that are currently in use, to some extent or another. One could be labeled the grade-equivalency model, which was used in the illustration at the beginning of this article, but there was none of this in the NALS. The grade-equivalency model is too vague to be of much use. Flippo, Hanes, and Cashen (1991) cited the match of reading levels to grade equivalencies as a misuse of reading tests.
Another model is the sub-skills model that is often linked to the basal approach to teaching reading. A basal series of elementary readers is likely to list skills and give a sequential plan for their teaching. This approach is not limited to elementary materials, but it is also used in some materials for developmental/remedial instruction. These subskills may be expressed in terms such as "Find the main idea sentence of this paragraph." While the NALS included some questions that seemed to call for the formation of inferences, which are often included in sub-skills lists, the survey did not seem to rely on the sub-skill model. The test developers did not seem to have written up a list of skills before they composed the open-ended test items.
The constructivist model assumes that the reader makes meaning from the text. This construction could proceed in unique, individual ways depending on the purpose for the reading. The NALS seemed to be considering some aspects of this model since the readers are asked to use prose and documents to integrate information and make inferences for certain contexts or purposes.
Some reading models are based on a schema approach. A schema is "... an abstract framework that organizes knowledge in memory ... (Nist & Mealey, p. 48). This model relies heavily on background information, and the NALS listed prior knowledge as one of the factors which readers had to utilize at level 5.
Now the whole language approach is contending with basal readers in some schools. One aspect of the whole language approach is the use of whole, authentic texts. The materials in the NALS were presented in the original format and had not been abridged or revised, so it corresponded to this aspect of the whole language approach.
While the NALS seemed to share fairly strong affinities with the constructivist approach, the schema model and whole language instruction, it also appeared to utilize the aptitude model that was developed and promoted for years by ETS, one of the developers of the test. An aptitude is an inherent ability. The NALS assumed that readers had the aptitudes to remember information, to use it and to disregard distracters. The aptitudes become more robust as the test levels rise and the passages become longer.
When developing a vocabulary for remedial reading in the community college, perhaps it would be helpful to consider the approach which the NALS took. The descriptions of the reading tasks were grounded in one or more compatible models of reading. A few key elements from these models were selected and a progressive mastery of competencies, from rudimentary to advanced, was sketched out. Terms which were linked to other models or systems, such as grade equivalencies, were not used. The characteristics of the text were listed and whole texts were used, not excerpts. For all the types of texts and competencies, open-ended questions were written to demonstrate mastery. Some of these approaches have been used at Community College of Philadelphia as we have considered reading, and perhaps we should continue.
Using the five levels of literacy in the NALS, it might be worthwhile to consider what distinguishes developmental/remedial reading from college reading. Barton and Lapointe (1995) pointed out that levels 1 and 2 were usually considered to show a lack of literacy abilities, and from 23% to 29% of the community college grads placed at these levels. It would probably not be too controversial to consider the first two levels of the NALS to be remedial/developmental and to have college reading start at level 3. This grouping would not inflate the numbers of developmental readers, and it would signal that, if the aptitudes at level 2 had been mastered, the college reader would be able to integrate information from different sources and make low-level inferences.
If the NALS results were plotted on a graph, community college graduates would tend to peak at level 3, in between high school grads who peak at about level 2 and baccalaureate graduates who peak at about level 4. This places the community college right in the middle, between high school and four-year colleges and universities. (SeeTables 1 and 2)
| Level | Level 1 | Level 2 | Level 3 | Level 4 | Level 5 |
|---|---|---|---|---|---|
| HighSchool | 16 | 36 | 37 | 10 | 1 |
| CommCollege | 4 | 19 | 41 | 32 | 4 |
| Four-Year | 4 | 11 | 35 | 40 | 10 |
| Level | Level 1 | Level 2 | Level 3 | Level 4 | Level 5 |
|---|---|---|---|---|---|
| HighSchool | 20 | 38 | 33 | 9 | 1 |
| CommCollege | 6 | 23 | 43 | 25 | 3 |
| Four-Year | 4 | 15 | 37 | 36 | 8 |
from Kirsch, Jungblut, Jenkins, & Kolstad, 1993, pp. 116-7
But this "mediocre" performance raises a question: Do community colleges want their graduates to occupy the middle ground? Shouldn't they be more like the other collegiate group, the 4-year graduates? In order to raise the level of community college reading, it might be proposed that level 3 be considered developmental. If this were the case, then the ranks of developmental students would increase greatly, to over half of the student body. It might mean, however, that the students who completed a developmental reading course and mastered the abilities associated with the NALS level 3 would be better equipped to handle the literacy demands which college would place on them. In the future, a greater percentage of the community college graduates might place higher on a test such as the NALS and appear to be more like baccalaureate graduates if the limits of developmental reading were changed.
Regardless of where the line between developmental and college reading is drawn, it probably would help to have a clear idea of what constitutes remedial/developmental reading and what constitutes "college" reading. The use of clear and appropriate terms may help college teaches as they discuss these distinctions.