Elba Herrero, Assistant Professor of Multicultural Education

 

The Language Assessment Battery (LAB) is used by numerous school districts to determine both native and English language reading skills of English-language learners. A number of public school teachers that I instruct have shared their concerns about the test. They report that many of their Spanish-dominant students have a great deal of difficulty with the LAB in both Spanish and English; that over the designated three years which students are given to exit the bilingual program, few score above the fortieth percentile in English. (#1)

After analyzing and discussing the LAB, many of my bilingual and ESL teachers arrived at the conclusion that the test has some general problems namely, that its cloze format(#2) and much of its content is culturally-biased. This paper discusses the cultural biases of the reading portion of the LAB to help explain why so many limited English proficient (LEP)students have had trouble with it.

First, a short discussion on the development of reading comprehension instruments. In 1973, Tuinman analyzed five popular reading tests and found that students could answer a significant number of questions correctly without recourse to the text because test items had surprisingly low passage dependency (Tuinman 1973-74). In 1982, Bensoussan noted that difficulty of passages did not match the difficulty of questions in recall protocols requesting explicit information (Bensoussan, 1982). Others focused on what makes test questions “easy” or “difficult.” Perkins and Miller (1984) suggest that test items which evoke paraphrasing, inferencing, and productive skills are more difficult than those requesting information derived directly from the text. Much of the research concludes that language proficiency is not considered in such second language (L2) reading comprehension tests.

The cloze format has been used widely to assess L2 reading proficiency because it has not yet been matched in strength with regard to degree of passage dependency. Cloze also validly measures and easily quantifies students’ ability to paraphrase and infer. These skills are considered by many reading specialists to be of a higher order than those associated with receptive measures such as recall.

Cloze protocols have a high degree of passage dependency (Berkoff, 1979). That is, test items are highly dependent on the text. Students must be able to pick out clues within the passage to supply correct answers.

My bilingual and ESL teachers examined various levels of the LAB’s reading portion for grades 6, 7, and 8. They noted the following: There is an average of 55 cloze items per reading section. Each reading passage begins with a complete (clozeless) sentence. Throughout the passage, however, there is at least one cloze item per sentence with as many as four per sentence. The ending sentences for each passage vary. One passage ends with three complete (clozeless) sentences while three passages end with one (clozeless) sentence. The length of the passage as well as the number of cloze items increases in order from the first to the last read. For example, in the last passage there are 13 sentences and 20 cloze items as compared to six sentences and five cloze items in the first passage.

Clearly, these LAB passages show that there is some sentence to passage interdependence. The text is sequenced so that clues carry over from one sentence to the next as well as from beginning to end. However, the issue of what the test measures is greatly dependent on the student’s ability to apply cloze techniques during the task. That is, the student who is able to transfer native language (L1) literacy skills to second language (L2) text is virtually excluded through this narrow assessment.

Most of the teachers I interviewed reported that their Spanish-dominant students showed better reading comprehension skills in both Spanish and English when responding to open-ended question formats than to cloze formats. Many concluded that the discrepancy in performance is in part due to the students’ lack of familiarity with cloze test strategies. That is, a number of their Spanish-dominant students fail the LAB because they are not familiar with the cloze format.

The general consensus was that these students find cloze reading passages unsettling because “the flow of the story” is interrupted by a great number of blanks. One teacher explains, “There are so many blanks in each story! My students can’t make sense of it.” Most agreed that these students need time and practice to learn how to do cloze.
We considered the question “When administered to students who are not familiar with cloze format, can it be said that essentially, the LAB is culturally biased?” Most of my teachers said “Yes”.

Research (Beach & Hynds, 1990; Bleich, 1978; Rosenblatt, 1978) indicates that readers interact in meaningful ways with text that calls up previous knowledge or experience that makes use of the reader’s schemata; i.e., what they already know, their interests, and their skills. With respect to L2 reading tests, Perkins and Jones (1985) suggest that passages which integrate previous knowledge with text-dependent items provide a more suitable assessment for evidence of text comprehension (p 151-152).

The teachers argued that a disproportionate number of their English learners fail the LAB because the test itself fails to engage them in meaningful reading activity; it excludes student schemata. They agreed that the LAB is culturally biased because much of its content is uninteresting to the students and tends to alienate rather than engage them in reading.

Each of the various levels of the test for grades 6, 7, and 8 contained four to six passages. All were expository in nature and covered three general categories: science, social studies, and technology. Indeed, we found that most of the content was not “culture-friendly.” For example, the Level III test was limited to the following topics: the roadrunner, the North American grizzly bear, Samuel Johnson’s dictionary, and helicopters. In addition, English-language expressions appeared frequently in the text. Specifically, “nature of the beast,” an idiom, is used to elaborate on the personality of the grizzly, a bear native only to the North American continent. While it does assess reading comprehension skills, it is also clear that the LAB measures knowledge of culture-specific content; i.e., North American wildlife, U.S. history and technology, and English-language idioms.

The students themselves seem to be the best critics of the test:
“Why can’t they give us something more interesting, like stories about people? I like to read about people’s lives.”
“Most of it is about things I’ve never heard of...animals I don’t know...things about this country.”
“I like science, but the way they use it here is stupid...not about interesting science...and there are so many blanks that you can’t really enjoy reading.”

To conclude, the teachers suggest that a more culturally appropriate test would use less (or no) cloze reading passages, include open-ended questions, and allow students to use their dominant language to respond to questions. They also agree that language testing specialists should consider student knowledge, interests, and experiences when selecting reading passages.

 

NOTES

(#1). A score above the fortieth percentile designates that a student is English proficient (EP), while those scoring below are labeled limited English proficient (LEP).

(#2). Paragraphs in cloze passages contain a number of blank spaces along with a column of choices for each blank. The text is rich in contextual clues. To make correct choices, readers must know how to apply knowledge of contextual clues within and across paragraphs.

 

REFERENCES

Beach, R., & Hynds, S. (1990). Research on the learning and teaching of literature: Selected bibliography (National Research Center on Literature Teaching and Learning Report Series R1). Albany, NY: University at Albany, State University of New York.

Bensoussan, U. (1982). Testing the test of advanced EFL reading comprehension: To what extent does the difficulty of a multiple choice comprehension test reflect the difficulty of the text? System, 10 (3), 285–290.

Berkoff, N. (1979). Reading skills in extended discourse in English as a foreign language. Journal of Research in Reading, 2 (2), 95–107.

Bleich, D. (1978). Subjective criticism. Baltimore: Johns Hopkins University Press.

Perkins, K., & Jones, B. (1985). Measuring passage contribution in ESL reading comprehension. TESOL Quarterly, 19 (1), 137–153.

Perkins, K. & Miller, L. (1984). Comparative analysis of English as a second language reading comprehension data. Language Testing, 1(1), 21–32.

Rosenblatt, L. M. (1978). The reader, the text, the poem. Carbondale, IL: Southern Illinois University Press.

Tuinman, J. (1973–74). Determining the passage dependency of comprehension questions in 5 major tests. Reading Research Quarterly, 9 (2), 206–223.