PUBLIC TESTIMONY OF KENJI HAKUTA

California State Board of Education

March 11, 1999

My name is Kenji Hakuta. I am Professor of Education at Stanford University. I have conducted research and published articles on the development of language minority students. Recently, I chaired a committee of the National Academy of Sciences that issued a research synthesis report on the education of language minority students. I am pleased to be here to testify against the proposed changes to Title 5, California Code of Regulations, amendments to Section 4306, regarding multiple criteria for reclassification of English Language Learners.

The first point I raise has to do with the removal of the requirement of objective testing for English oral proficiency. This proposal is problematic on the grounds of both policy and practice.

From a policy perspective, this is an unfortunate reversal of the national trend toward objective assessment of student progress toward meeting performance standards. I am happy to echo what Don Iglesias, Assistant Superintendent of Instruction of the Santa Cruz City Schools and Member of the Board of the Association of California School Administrators wrote in his letter to your Board: "I believe the deletion of all specific language relating to criteria for reclassification will make it impossible for the state to gather any meaningful data relating to the number of English Language Learners who have become proficient in English. It is critical that school districts be required to report the progress of English Language Learners toward reclassification within their district using the same assessments and criteria as other districts throughout the state. Otherwise, there will be no way to assess the statewide progress of English Language Learners on a consistent, measurable basis through California."

From a practical perspective, subjective evaluation of student English proficiency is incomplete and fallible. Research on the relationship between subjective judgment and objective assessment of language proficiency show generally positive but far from perfect correlations. I can illustrate this through Exhibit I, which shows a variety of scenarios of how objective assessments might be predicted by subjective judgments. Panel A illustrates the common person’s view of what it might mean to say that subjective judgments are as good as, and substitutable for objective assessment. Visual inspection of the graph will show that the points line up quite well, such that knowing the subjective score would be an accurate proxy for the objective score. But as you move from Panel A to B, and down on to F, you will notice that the correlation becomes less than ideal. At F, reasonable gamblers would not bet much money in guessing the objective score even if they knew the subjective score! Based on research, we can say that the real relationship between objective and subjective assessments of English proficiency fall somewhere between C and E (or, statistically speaking, between r=.5 and r=.7). Another way of putting it is that there is a lot of scatter between the two, and one would certainly not want to place the fate of individual students solely on subjective judgments. Research on subjective ratings also show that the quality of the ratings are affected by factors such as the rater’s attitudes and the training they have received, factors that I know will vary a lot from district to district. In sum, subjective evaluation is a very poor substitute for objective tests. The Board would be setting very low standards of accountability by not requiring objective testing.

My second point is to strongly reject the assumption that we only need to worry about the English oral proficiency development of English Language Learners. The proposed change eliminates the requirement to look at other measures of their academic competitiveness, including reading and writing in English. In my opinion, that is a big mistake. I would like to turn your attention to Exhibit II, which is data I have collected from a district in the Bay Area with a high rate of success in educating their English Language Learners. These data show, for example, that by 6^th grade, about 80 percent of the students who started their schooling as limited English proficient were redesignated, using the current criteria. The district uses objective measures of oral English, reading, and writing. As we can see from their growth profiles, reading and writing take considerably longer to develop (and the happy point is, that they do develop!) than oral English. The reading criterion is set at equivalent to the 30^th percentile in norm referenced tests of English speakers – at that point, they are at least beginning to be competitive with their English-speaking peers. Thus, the research clearly shows a need to take into account not just English oral proficiency, but other criteria as well that are predictive of their academic competitiveness.

I appreciate the time and attention you have given me. I would like to close by saying that I am a strong believer in accountability. A good part of my public service time in the last 7 years has been expended toward arguing for their inclusion in the standards-based reform movement in the various content areas. I had always assumed that a big part of this accountability would be in their English language development, and not just in academic content. Thus, I am especially distressed that you are seriously considering a proposal to eradicate any possibility of having a state-wide system to account for their English language development. For anyone who seriously believes in academic standards, this is a major step in the wrong direction; I hope that the Board chooses to exercise leadership in a positive direction on behalf of the English Language Learners of this state.

Kenji Hakuta

School of Education

Stanford University

Stanford, CA. 94305

(650) 725-7454

hakuta@leland.stanford.edu

http://www.stanford.edu/~hakuta