Letters to the Editor
To the Editor:
Michael H. Kean makes a number of statements in his Commentary, "Getting It Right: Authentic Assessments and the True Multiple-Measures Approach'' (Oct. 6, 1993), that call for comment.
1. Norm-referenced, multiple-choice tests are "valid, reliable, and fair.'' The question is what is the "norm.'' If it is getting a good grade on a multiple-choice test, this may have some validity. If it is driving a racing car, or swimming, or selling automobiles, or making electrical repairs, or speaking French, I have my doubts.
2. Performance assessment is a "newly resurrected format.'' Performance assessment is not only the type of assessment that goes back to prehistoric time, it has been in continuous use in every functioning organization throughout time. It has been recognized as the only valid assessment instrument in situations as different as passing puberty rites in Native American tribes to selecting the C.E.O.'s of major international companies in 1993.
3. "Multiple measures are the only sure way to get valid, reliable, and fair results.'' This is the most questionable of all of Mr. Kean's statements. It must be obvious that adding invalid scores to a valid score does not increase validity. It might be more valid if Mr. Kean added parents' income to a valid assessment of a student's performance rather than scores of multiple-choice tests of information.
4. " ... [I]t is precisely here that the policy debate over a single measure--norm-referenced, multiple-choice tests versus performance assessments--begins to sound silly.'' I don't quite see the precise point at which Mr. Kean sees "silliness'' coming into the debate, but the differences are far from "silly''; they not only affect assessment of school performance, they are important for job selection and implementation of equal-opportunity laws.
5. Some arguments can be made for well-constructed multiple-choice tests for assessment in limited knowledge areas that take into account characteristics of the test population and the use to be made of the test results. But few of us who have made extensive use of assessment instruments and have reflected and been concerned over the validity of multiple-choice and similar paper-and-pencil tests are as sanguine as Mr. Kean about the general applicability of these instruments.
New York, N.Y.
The writer is a retired chairman of the City Civil Service Commission in New York City and was personnel director for the City of New York.
To the Editor:
Denis P. Doyle's response to my Sept. 29, 1993, Commentary, "George Will's Urban Legend,'' marks a historic moment (Letters, Oct.13, 1993). Surely it is the first time he has ever accepted on faith a statistic uttered by a National Education Association official. (Oddly enough, in a letter to me, George Will accepted it also). The N.E.A.'s president, Keith Geiger, erred, however, and in a conversation that occurred while I was tracking down Mr. Will's error, referred me to Mr. Doyle's analysis.
Mr. Doyle's protracted protestation is quite bizarre. Of my essay, he says that "again and again he quotes statewide figures using states with historically low private school enrollments.'' He accuses me of ignoring "the central point of our research--what urban teachers do.'' But that is not the central point of the research (I invite readers to obtain copies of the unpublished piece from Mr. Doyle and make their own judgments).
Mr. Will alleged a national statistic existed and it was the existence of that statistic I sought to dispel. Mr. Doyle's and Mr. Hartle's manuscript, the only such analysis to date, dispels it utterly. While Mr. Doyle accuses me of selecting states with low private school enrollments, I used only those states that Messrs. Doyle and Hartle provided in their article, and used all of them.
I have no real quarrel with Mr. Doyle over the higher percentages of teachers using private schools in large central-city areas and don't understand why he feels obliged to attack me on this point. I have written about the terrible conditions of many city schools and the poor performance of students therein. Indeed, my article to which Mr. Doyle objects wonders "why the figure is not 50 percent or higher.''
It is worth noting that in the case of Chicago, about which Mr. Doyle makes so much, fully 87 percent of the public school teachers with children in private schools had them in parochial schools, suggesting religion plays a significant role in school choice. Non-academic concerns are also important. Said the Chicago Reporter, "Many teachers believe parochial schools offer more discipline and serve as havens from violence and gangs.''
Gerald W. Bracey
To the Editor:
Will the educational-research and -measurement professionals ever admit that they can measure anything? The study described in "Yet Another Report Assails NAEP Assessment Methods'' (Sept. 22, 1993), continues their relentless and increasingly strident attempts to discredit any attempt to measure how well U.S. students are performing academically.
Over and over, they've told America how weak our tests are, whether teacher-made tests, standardized achievement tests, the old S.A.T., the new S.A.T., the old NAEP, and now the new NAEP. Some (but thankfully not all) of them tell us that only portfolio assessment will give us the real picture. The problem is that this method is complicated, costly, hard to summarize, and hard for parents to understand and for decisionmakers to use. But maybe that's O.K., because then no one will know how poorly our education system is doing. That's one way to avoid negative stories.
Let's remember the questions parents and the public want answered! How well do our children read, write, compute, and think: 1. compared to a reasonable, understandable, objective standard? and 2. compared to their peers in the next classroom, school, town, state, and country?
Why is it so hard to answer those questions if one is not trying to hide something?
Hyde Park, Mass.