|
12/25 |
2005/4/28-30 [Academia/GradSchool] UID:37402 Activity:insanely high |
4/28 Why is there so much opposition from liberals to a high school exit exam? Supposedly, 83% of kids have passed it. I know it is bad for self-esteem when a kid is a fuckup, but maybe failing the exam will help that lower fifth realize they need to work harder. The funniest comment I heard was the exit exams will hurt graduation rates. Yeah, I guess they will. If a kid can't pass an easy exam they have been preparing 3 years for then maybe they shouldn't graduate until they do. Why are these people teaching our children?! \_ Uhm, why would you need an exit exam in the first place? Instead of making kids pass an exit exam before graduation, we should just make the classes harder. We don't need an exit exam to get a degree from Berkeley, we just need to pass all of our classes. -moderate \_ Lots of universities do have exit exams. \_ Interesting, which ones? What do they test on? \_ Well, for example, UC Santa Cruz. It's a department-level requirement rather than a university one. The exam covers general knowledge, but a list of study materials is provided ahead of time. \_ How do you enforce that all classes meet a certain threshold of quality and grading levels are meaningful across schools? Hell, forget K-12. How do you do that at the college level, so that graduates of 'fraud with a A- GPA knows as much as a Cal grad? \_ Uhm, give more money to schools so that they can hire better people to A) Administrate them and B) Teach the courses in the K-12 level? As for the college level, in order to pass a compsci class you need to at least know how to program to a minimal level of proficiency to pass, at least at Cal. As for other schools, well, the ones that don't matter and have easy classes have bad reps and nobody cares about them. The ones that have easy classes (like Stanford and Harvard) have a built-in selection mechanism to ensure only bright people get in or their parents have enough money to ensure that their kids have a minimal threshold of education to start off with, so society takes care of them. Anyway, if you want better HS graduates what you need to do is change the culture and the curriculum. I don't get how a standardized test achieves either. In a sense, we already have a standardized exit exams for college bound high school students, the SAT, so an additional test of that sort would be redundant. If you want better schools you first need to get better administrators and second you need to get better teachers and third you need to get better infrastructure. All of the above require lots of money, which is why prep schools and other private institutions blow public schools out of the water in general. Also, if you take a look at school districts that are part self funding like the one in Contra Costa County you see a world of difference. It's actually pretty simple, want better schools? Increase school funding. The question is, do you want to pay? \_ The money arguement is provably bullshit. If you look at the spending per student at different schools around the state(which I just did for my home state), you'll see shitty, crime-ridden inner city schools often spend more money per student than the suburban schools, which in my state gave the prep schools a run for their money in terms of test scores and admission to top colleges. \_ You're forgetting the amount of money that schools raise on their own through fundraisers plus the infrastructure that parents contribute to the schools in suburban communities (i.e. the PTA in suburbs is MUCH more active than in inner cities and a LOT more people are privately funding things behind the scenes). The reason why inner city schools get more money is because they have to deal with a lot more problems, i.e. security. A suburban public high school in a decent neighborhood will equate to an inner city prep school. You can't simply look at the raw numbers that the state provides you. Kids in suburbs cost a lot more to raise per capita than kids in inner cities. Parents have the means and are willing to spend that kind of money on kids in suburbs, but parents in inner cities do not. Again, it's just a function of money. If you want higher test scores and smarter kids, be willing to spend the money to upgrade their environment. Again, are you willing to pay? \_ As usual, you miss the point. The point is that *no* amount of money will solve the problem of bad inner city schools. The kids in the public school I went to got higher test scores because they were in a culture that encouraged academic performance. Most kids I knew were read to by their parrents before they even got to school, which gives us an advantage that no amount of spending can make up for. I'm not proposing a solution, I'm just saying that lack of money is simply not the problem. Calling greater parental involvement in schools "infrastructure", as though you can just add that to a town budget is just plain stupid. \_ Uhm, no, you miss the point of my previous post. I never said that just adding in money will result in better schools. I said that you need to change the administration and the teachers and the infrastructure of schools to get higher test scores. In order to do that you have to pour in resources into the schools, which basically equates to money. Obviously pouring money into an existing infrastructure that doesn't work will not work. I'm saying you have to change the system. Changing the system requires a lot of money and a lot of political will power. The question is, are you willing to spend time/money on this problem? It's also naive of you to think that good schools can't make a difference in a young person's life. Ever see the move Stand and Deliver? It's a true story. I actually was close to the district that Garfield High was in. Thinking that Latino or Black kids can't learn Calculus was wrong and was racist. Give the schools the right infrastructure, and you can turn out inner city kids that can ace the AP Calc exam. Case closed. \_ I think the point here is that the inner city schools receive about the same funding, so why does their infrastructure not work and yet in suburban schools it does? Sometimes this happens at different campuses in the same school district (e.g. LA Unified). \_ Grade inflation is rampant throughout higher ed, including at Berkeley. http://ls.berkeley.edu/undergrad/colloquia/04-11.html. 50% of Berkeley undergraduate grades are A's, 35% are B's, and less than 5% are D's or F's. -tom B's, and less than 5% were D's or F's. -tom \_ The grade distribution never looked like this for the classes I took (mostly math). Didn't the Physics Dept. mandate a bell curve with 50% at C+? If you scored the mean on a test you didn't usually get an A (or even B+). That might be different in, say, anthropology or classics where I got A's without really trying. \_ They note that there are differences across disciplines. -tom \_ Indeed. They do mention that in Physics just 15-20% of students earn A's. "The Physics department began monitoring lower division GPAs at one point when it was discovered that instructors were giving mostly B's." \_ Robert Holub hosted the event. Any relationship to tom? \_ No. -tom \_ Definitely good news for people who had a hard time graduating, like t** and p** \_ "A significant increase in the GPA occurred during the Vietnam War when students received a draft deferment if they remained in good academic standing." Who says draft is a bad thing now? :) \_ The SAT is for entrance to college, not graduation from high school. The idea behind an exit exam is that it gives more value to a HS diploma and is also a metric that schools/teachers can use to see how well the students are doing at all levels, not just college-bound students. \_ So my question to you is, why not just make the SAT a metric that employers can utilize to gauge students? If you are so concerned about gauging the metric of the average H.S. grad, make them take an existing standardized test. Why should the state take up the burden of creating another standardized test? Anyway, companies like UPS already utilize standardized testing to screen applicants, making the whole point moot. \_ It's not for the employers to 'screen', but a metric for the schools. They could use the SAT, but they have no control over it. \_ Exactly... you say bad colleges get bad reputations. That doesn't work at the high school level. They are providing public services and need to be held to some minimum standard. They can never be equal but there's no way to enforce "making classes harder" other than having harder standardized tests. \_ How do you know if schools are getting better or worse without testing? What metric do you propose? How does that metric compare to direct testing of the output? \_ Perhaps you should give your examples of "liberal opposition". The progressive take on this is not necessarily that exit exams are bad, but that standardization with negative reinforcement is not the way to help an ailing school district. \_ I gave my examples. Bad for self-esteem and lower graduation rates. Obviously not all liberals feel this way, but all of the opposition is indeed liberal. \_ I meant examples of liberal opposition. Not the factors to which you say they object. How 'bout an article or two? Transcript to a news show? \_ Find them yourself. It's easy enough. It's in the headlines right now. \_ It's your argument, man. If you can't support your own assertions with evidence, or show how you reached your conclusion aside from faulty logic or opinion, then you're in serious danger of being labelled (quite justifiably) a troll. If you can't even support your own argument with evidence, why the fuck should anyone else? \_ I am assuming I am debating with informed individuals. If you are not informed then it is easy to become so. Use Google. Otherwise, I don't have time to search for and post links. This is in the headlines. It's like asking for a link to who Schiavo is. Read the Chronicle, which said: "Leaders of the teachers unions are adamantly opposed to the exam, as are groups representing the minority students with the lowest pass rates." I'm not a frickin' newspaper. Read one once in a while. \_ why do conservatives support stealing money from public schools to give to rich families? \_ Because public schools apparently aren't worth shit according to the standardized tests. However, most wealthy conservatives not only put their kids in private schools, but contribute to public schools as well. \_ So you're saying conservatives don't care about the non-wealthy (which is most of them) conservatives? \_ They probably don't, however which part of "contribute to public schools as well" did you miss? Whether they care or not, they are helping anyway. \_ If by "liberals" you mean "teachers", then I can tell you it's because they look at education as a continuing process, while those who implement std'ized tests look at it as a race with an endpoint. In a perfect world, they would have small enough class sizes that they could give the necessary attention to each student. In their world, the number of students who passed whatever test was set in front of them would be much higher than 83%. As it is, it's a deadline put upon a system with limited resources. When budget cuts are linked to poor performance on these tests, it creates an incentive to "teach to the test", to the detriment of actual education. \_ This is a load of shit. So "teach to the test" then. At least we're sure they are teaching something, as opposed to now where kids get by w/o learning shit. \_ Dim, you're a load of shit, as you make clear any time you post here. Talk to a teacher about this sometime. Pick one you respect. You'll be surprised how much you don't understand. \_ I've talked to a lot of teachers and I think they are mostly afraid of facing the reality of their situation. No one said teaching is easy. I respect that. However, standardized testing is not supposed to be a panacea or a way for kids to learn more. It's just a metric. If you want to propose something akin to a thesis to graduate high school I am all for that, but personally a simple test seems a lot easier for the teachers if less accurate. \_ So you've talked to the teachers and dismissed what they've said out of hand. Grow up. \_ Yes, it is a sign of immaturity that I do not take what teachers say at face value. Heck, my sister-in-law is a teacher. I find their arguments lacking. Most of it is touchy-feely bullshit about catering to the lowest common denominator. These are the same people who don't want to be paid for their performance. I can't even imagine having a job where my performance wasn't tied to my raises. Why would I want to make the same as someone who does less than I do? For teachers somehow this is OK. \_ Yes, not everyone in the world has the exact same worldview as you. That does not mean that they are "wrong" and you are "right." \_ I have the freedom to label them 'stupid'. Many of them are booksmart, but have no clue how anything works. It comes from being around 8 year olds all day. Not wanting a raise for performance & not wanting the underperforming teacher in the next room replaced is idiotic in every worldview but theirs. If schools were run by businesspeople instead of 'educators' more shit would get done. In fact, this is closer to how private schools work. \_ I'm amused that you think you have a clue about how things work. \_ I get a feeling you are easily amused. I'm saddened that you think I don't have a clue. \_ This is a load of shit. How is 'teaching to the test' equivalent to an education in any rational sense? Memorizing a set of answers without any context or any ability to apply that knowledge isn't education. \_ Our education system, including college, is the best in the world. We must be doing something right. \_ Nah - I think our collegiate educational system is the best in the world, but our lower level educational system(s) are desperately in need of attention. \_ Your claim that high school graduates are the worst in the world is hard to reconcile with the fact that our college graduates are the best. \_ Uhm, ehr? I never made the claim that our HS grads are the worst in world. You should reread my previous post, dude. the world. You should reread my previous post, dude. \_ Maybe most of our college grads are foreigners (this is certainly true in the postgraduate level). \_ Really basic math skills, history recital, and English language ability should be easily testable. And yes, a lot of kids would fail that. And if they do then they aren't ready to graduate. Obviously we expect that an education should have been provided along the way, but a basic test can at least stop blindly pushing kids through a system without even meeting the most basic of educations. Ideally basic tests should be given in earlier grades to catch problems earlier. Kids in a certain grade should be expected to have a certain skill level. Smaller class sizes are good to a point, but only the kid and his parents really have the ability to make sure a kid learns actual skills, and not sit around in no-pressure environments where everything is the right answer. \_ Err, I'm not taking a stand for or against the notion of an exit exam. Reread, please. I'm objecting to the fuck-stupid notion that teachers prepping students to pass one very basic test is in any way meaningful. \_ The problem here is that kids aren't passing the tests EVEN WHEN the teachers teach to the test. This implies that when they are not teaching to the test the results are about as dismal. So teach to the test and get 98% of the kids to pass and then worry about if it is meaningful. \_ And yet our fully contextual students get clobbered annually in achievement tests by students in countries big on memorization. And we lament when our colleges and grad schools couldn't import more of those memorization drones. Amazing. \_ 'Fully contextual'? Are you nuts? The problem has more to do with crappy quality of education and (depending on who you speak with) a bloated administration that soaks up any money thrown at it. If you think that memorizing answers for one test is going to magically fix everything, you are so deluded or ignorant it makes my teeth ache. \_ college admissions use mostly GPA and SAT to select students. even the private ones who are free of government garbage. so the free market thinks standardized tests are a good indicator of academic prowess for their student body that they want to be the best possible to generate good alumni etc. \_ 'The free market'? LOL. Thanks. That actually made my day. \_ Do they make the tests available ahead of time? Or the questions? If not, how do you memorize the answers? Or do they make a study guide available? Go ahead, memorizing that would be a good start. \_ Yes, actually they do make the tests available ahead of time. That's a rather large part of the origin of this debate. That's why you need tests that can't be taught to. \_ URL please. Giving out tests early seems stupid enough to require substantiation. If all they give out is a study guide, then teaching to that seems quite reasonable. It would be even better if they called it a "study guide" but it's really a "text book". \_ That rocks. \_ In fact, that seems to be exactly the case. "Teaching to the test" means teaching the subjects known to be in the test, rather than teaching the questions (or answers as you claimed). See http://csua.org/u/bwm for example. I'm still awaiting proof of your claim that tests are given out early. \_ Find them yourself. It's easy enough. You know *nudge,nudge* google? Yahoo? \_ I can't find any. Of course, it's impossible for me to prove that it doesn't exists, prove that it doesn't exist, hence my plea of a positive instance where the test was given out early, since you made the claim. You can't find one either, huh? \_ "The abscence of evidence, \_ "The absence of evidence, is not the evidence of abscence." absence." \_ One would think that you made the claim with a particular example in mind... Otherwise why would you make the claim in the first place? So you're saying you made the original claim of tests being handed out early to students *without* any basis at all? \_ No, I was just making fun of Rumsfeld for saying that during the WMD debate. \_ So where is the reference to tests being given out early? Again, one assumes you have made the claim with some basis in fact. \_ I am not the same guy who made that claim. I am some other jackass. \_ Your assumption that he ever looked may be somewhat specious. \_ I listened to an interview on KCBS radio. Teaching to the test is big. The interviewer asked, "Is that because the teachers are teaching to the test?" The researcher said, "No, we use an adaptive test that cannot be taught to." Interviewer: "How about this other measurement that declined? Could that show teaching to the test?" Researcher: "No, it's impossible to teach to our test. Our data don't show why that other measure declined." 3 questions later, the interviewer asked: "Does this show the teachers are teaching to the test?" Researcher: "We don't know yet. You can say that if you want, but we haven't done the analysis yet." End of interview, interviewer: "Apparently teaching to the test is causing an improvement in test scores." \_ I'm a liberal and support HS Exit Exams. \_ I'm a liberal, and I'd support them if they reflected the results of some sort of organized curriculum of basic material. \_ I'm a liberal and I support gays, lesbians, and welfare. \_ You would understand the opposition if you looked at the composition of the 17% who fail the test. \_ Yes! I forgot this argument. It's racist! Tests are racist plots invented to keep minorities down! Oh, except Asians \_ Standardized tests are culturally biased in favor of those cultures that value education. \_ Have you actually considered the possibily that standardized tests have cultural bias? Nah, probably not. You sound too smug and arrogant to ever consider the possibility that your assumptions might be incorrect. \_ This arguement is laughable, and makes me ashamed to call myself a "liberal", since it seems to be only liberals who actually believe it. Maybe you can explain what Irish, Jewish, Indian, Japanese, Korean and Scottish immigrants have in common that somehow makes tests biased in favor of all of them. I'm pretty sure that if the tests were somehow geared towards people who were raised in, say, a Japanese household I would have failed. And if they had a "jewish" bias, I'm sure the same would be true of most asians who also kick ass on the tests. I'll say it again: standardized tests are culturally biased for exactly one cultural trait: valuing education. This is why affirmative action for higher education makes sense. \_ Are you the same guy who mocked the the notion that there was bias in the tests? You are talking out of both sides of your mouth, if you are. And yes, I can explain how it might be so, but I will not bother wasting my time with someone who is mind is so obviously already made up. \_ Fucktards like you are the reason we keep losing elections. Maybe you should take your giant brain over to the republican party. \_ Old jungle saying: You can lead a girl to Vassar but you can't make her think. \_ Apparently they're culturally biased for Asians. \_ Is that what you think? Perhaps they are simply biased toward the wealthy: http://www.america-tomorrow.com/ati/gb80211.htm \_ Asians (Jews, too) didn't always have it so good. I am not sure it's genetics, but it certainly is cultural. They are wealthy because they worked hard and studied hard. You have your cause and effect mixed up. \_ At least you are thinking, unlike Mr. Your Argument Is Laughable fellow above. There is probably some kind of virtuous cycle that is taking place. Whether you want to call this evidence of bias or not is up to do. Here is a great paper by a Harvard researcher talking about cultural and language bias: http://gseweb.harvard.edu/hepg/freedle.html |
12/25 |
|
ls.berkeley.edu/undergrad/colloquia/04-11.html Click to jump to section links of this category (if any) or continue f or page contents Undergraduate Division L & S Colloquium on Undergraduate Education L&S Colloquium Tackles Grades and Grading Philosophies By Susan Hagstrom January 19, 2005 "91% to Graduate with Honors (at Harvard)" "Grade Inflation: A Crisis in College Education" "Grade Inflation: It's Not Just an Issue for the Ivy League" Recent eye-catching headlines on grade inflation prompted the Undergradua te Division of the College of Letters and Science to host a colloquium o n the topic of grades and grading philosophies. On November 22, 2004, a panel of four distinguished administrators and fa culty members, and an audience of faculty, staff, and students, gathered to discuss the meaty issues surrounding grades and grading philosophies . Dean of the L&S Undergraduate Division Robert C Holub moderated the even t and opened the discussion with his perception that no topic is more im portant to students while they are undergraduates and less important aft er graduation from college. The four panelists then presented thoughtful , provocative perspectives on the value and purposes of grading along wi th some specific views about what's working, what's not, and what the in stitution could do differently. In a reference to Garrison Keillor's fictional small town, panelist Dr. D ennis Hengstler of the Office of Planning and Analysis described the phe nomenon of grade inflation as the "Lake Wobegon effect", one in which al l of the students are "above average." Giving a historical perspective, Hengstler noted that in prior years a grade of "C" was considered "Avera ge" and is now defined as "Fair." Hengstler set the stage for his fellow panelists by providing statistics and facts about grades at UC Berkeley and elsewhere: * In the late 1950's, the average cumulative GPA for Berkeley undergrad uates was 250 and has increased to approximately 325. A significant increase in the GPA occurred during the Vietnam War when students received a draft deferment if they remained in good academic standing. These findings parallel recent data (spring 2004) from the National Survey of Student Engagement: 40% of students say they earn mostly A's, with 41 percent reporting that they earn mostly B's. An exception in recent years occurred when the SAT total score decreased between 1998 and 2001 as it was deemphasized in the UC admissions process. Such variations are a national phenomenon and are not specific to the Berkeley campus. Princeton has declared that "A's shall account for less than 35% of the grades in any department." "No one, not even faculty who h ave been here for 20 years or more, can accurately interpret a transcrip t," stated Rine. "Yet we try to serve our students, often writing letter s of recommendation, based upon the very data on those transcripts that we cannot accurately interpret." Rine noted the variance between policy and practice in citing a 1976 Berk eley Academic Senate recommendation that the average grade awarded by th e instructor in a course be recorded on the student's transcript along w ith the class size and the grade he or she has earned. The Academic Sena te also stated at that time that "it seems to us that we should attempt to return to the traditional distribution where grades A and B recognize honor work in undergraduate courses and should be awarded to fewer than half the students on average." Rine described the shock he felt during his three years on the Committee on Teaching from roughly 1998 to 2000 when he reviewed teaching records for large undergraduate classes, with more than 100 students, in which n o one got less than an A-, year after year. At the time, Rine asked Asso ciate Registrar Walter Wong to assemble some data looking at upper divis ion and lower division grading in the physical sciences, biological scie nces, social sciences, humanities and engineering, so that he could dist inguish trends from anecdotal exceptions. "The p hysical sciences and engineering had rigorous grading standards roughly in line with the recommendations from 1976," stated Rine, "while the hum anities and social sciences in many classes had all but given up on grad es below a B, and in many courses below an A-, and the biological scienc es had no consistent pattern." Rine gave data from his own discipline to illustrate: in the lower di vision, the average Bio 1A GPA is 248. At the other extreme, in MCB61 t he average GPA is 328, nearly a full grade point higher. At the upper d ivision, MCB 100 has a rigorous 257 GPA, whereas MCB130 L has a 350 GP A, with 62% of the grades being A's. Referring to the 1976 recommendation, Rine outlined two possible benefits of recording the class size and GPA on the student's transcript. First, a student could tell whether he or she was adequately measured against the other students in that class, and hence would have some feeling for whether he or she has talent in that field, as well as an understanding of how much work was required to achieve what level of distinction. Seco nd, faculty evaluating transcripts "would be able to see whether there i s clear evidence of distinction in rigorously graded classes, or whether the grades are ambiguous, in which case we would be better off emphasiz ing other aspects of the record in preparing our evaluations of that stu dent." "The student turns to the last page of the paper and either congra tulates himself or condemns the instructor. When asked to revise a paper, most students use the instructor's comme nts to simply 'correct' their work." Now, McQuade chooses not to put grades on papers, maintaining that grades are a serious distraction from the pedagogy. This approach keeps the re sponsibility for making progress where it belongswith the student. McQu ade believes that his first obligation as a teacher is to create conditi ons where students can learn. He asks students to make an ethical commit ment to submit work that is "ready-to-be-read." The initial readings are done in small groups of students. At the end of the term, McQuade also asks students to write a detailed analysis of the work they have done du ring the semester, summarizing the nature and extent of the specific pro gress they've made and recommending a final grade. If there's a discrepa ncy of more than a half grade (B / A-), he meets with the student to rev iew and reconcile the difference. He finds that the vast majority of stu dents undervalue their work. McQuade observed that students often are expected to pretend that they kn ow how to do something before they've had the time to establish mastery over the skill through practicing it. So, too, he noted that one goal of teaching is to make oneself obsolete by the end of the semester. Bob Jacobsen distinguishes between the grade rec orded on the transcript and the mastery of course material. For Jacobsen , an A should indicate a very good level of understanding of the course content. Jacobsen states that just 15 to 20% of his students earn A's be cause he hasn't figured out how to teach well enough so that they all ma ster the material. "Student's worry about preparing for impacted majors," reflects Jacobsen. "They don't worry about whether they are adequately prepared for the ne xt course." A C grade should indicate that a student is prepared for tha t next course, but may have to struggle with some of the prerequisite ba ckground. The Physics department began monitoring lower division GPAs at one point when it was discovered that instructors were giving mostly B's. A facult y committee then mandated that faculty give A's and C's or explain why. This worked for a while but now, Jacobsen notes, the department is once again drifting toward giving mostly B's in lower division. "I'm not willing to say that 17% of students should automatically receive A's and B's. But I am willing to say that 25% should get D's and F's if they earned them." L & S colloquia, which take place once or twice each semester, pro vide opportunities to learn about and discuss the overarching issues aff ecting undergraduate education at U C Berkeley. |
www.america-tomorrow.com/ati/gb80211.htm THE SECRET TO THOSE HIGH ASIAN TEST SCORES: AFFLUENT, WELL EDUCATED DADDI ES, MOMMIES, TOO Gerald W Bracey A principal I know from Brooklyn once joked that when he was a kid, all o f the students knew that Jewish kids had a "smart gene." Now, he said, t he students in this same Brooklyn neighborhood know with equal certainty that the smart gene resides in Asian brains. Indeed, the performance of Asian-American students on tests has been the subject of much admiratio n and why-cant-the-rest-of-you-be-like-this commentary. In a 1992 international study of mathematics, Asian kids in this country ev en outscored students in the top two nations, Taiwan and Korea. Now comes a report from Educational Testing Service that reveals some mor e mundane reasons for high Asian-American test scores: contrary to the s tereotyped images, Asian families are not huddled in tiny apartments in various "Chinatown" slums. Asian students live in the suburbs with paren ts who are considerably more affluent and better educated than the natio n as a whole. If we put them all into a single district, in terms of edu cation and income, it would look a lot like Fairfax County, Grosse Point e or Cherry Creek. In the year of the study, the median family income for Asians in this cou ntry was $41,251. For the nation as a whole, the average family that yea r took in $32,142. The difference is the more remarkable when one consid ers that the figure for Asians includes Cambodians (income, $18,126), La otians ($23,101), and Hmongs ($14,327). Vietnamese, though, had an avera ge income better than the national average, $33,909. When we look at the educational attainments of our Asian immigrants, the figures are staggering. Nationally, 31% of the po pulation are alumni of some institution of higher education. The people coming here from Asia are not just those beckoned by the inscription on the Statue of Libertys base. The figures for other groups, while not matching the South Asians, are al so impressive: Japanese fathers with college degrees, 53%, mothers 42%; F ilipino fathers, 38%, mothers 41% (Filipinos are the largest Asian immig rant group, by the way); Only mothers eman ating from Southeast Asia fall well below the national average, 12%. From these figures alone, someone could argue that this proves that Asian s have a smart gene. But South Asians, with 87% college degrees and Fili pinos with 38% are not among the groups generally presented when people speak of high scoring Asian students. Every study that has ever looked at the relationship between parental edu cational level or parental income and their kids test scores has found a simple correlation: more money equals higher test scores; This income-education-test-score ladder sho ws up on the test results of Asian students. For the best educated, weal thiest group, South Asians, 79% of their offspring score above the 50th percentile in reading, 84% above the 50th percentile in math (nationally , by definition, 50% of students score above the 50th percentile, 50% be low). The percentages decline with declining income and education until, for th e least educated, least affluent group, Southeast Asians, students are a ctually below average: Some 43% of Southeast Asian students scored above the 50th percentile in math, only 32% in reading. Little wonder, then, that only 65% of Southeast Asians aspire to earn a college degree compar ed to 95% of South Asians and 90% of most other groups. Actually, the sc ores for Southeast Asians are somewhat inflated because the report also notes that in high schools with high concentrations of Southeast Asians, the dropout rates hover around 50% (that sound is the crash of another stereotype). The really low scorers have left, taking their low scores w ith them. It would appear that in addition to welcoming the worlds tired, poor, hud dled masses, we have also been aggressively selecting educated, skilled immigrants from the Asian continent. The ETS report was designed to disp el the notion that "Asian" is a meaningful label for such diverse cultur es as Japan, Vietnam and India. It has the incidental merit of sending t he "smart gene" up in a puff of affluence and schooling. Gerald W Bracey is an independent educational researcher in Alexandria, VA and author of the annual "Bracey Report on the Condition of Public Ed ucation (Phi Delta Kappan, October) and of Setting the Record Straight: Responses to Misconceptions About Public Education in the United States. |
gseweb.harvard.edu/hepg/freedle.html com Abstract The SAT has been shown to be both culturally and statistically biased aga inst African Americans, Hispanic Americans, and Asian Americans. In this article, Roy Freedle argues for a corrective scoring method, the Revise d-SAT (R-SAT), to address the nonrandom ethnic test bias patterns found in the SAT. The R-SAT, which scores only the "hard" items on the test, i s shown to reduce the mean-score difference between African American and White SAT test-takers by one-third. Further, the R-SAT shows an increas e in SAT verbal scores by as much as 200 to 300 points for individual mi nority test-takers. Freedle also argues that low-income White examinees benefit from the revised score as well. He develops several cognitive an d cultural hypotheses to explain the ethnic regularities in responses to various test items. But, in practice, this early promise has not been fulfilled, especially for minority groups whose mean test performance has departed significantly from White mainstream test-takers . Over the past several decades, the search for a more equitable ethnic representation in our nation's select colleges led to the adoption of af firmative action, a policy that is increasingly under attack (Lemann, 19 99). The recent erosion of affirmative action increases the need for oth er means of identifying promising minority students for admission into o ur system of higher education. This article suggests one avenue for solv ing this problem. The chief purpose of this article is to present a new method of scoring t he SAT (called the Revised-SAT, or R-SAT) that will greatly increase the number of high-scoring minority individuals. This new scoring method co rrects for two types of potential test bias: cultural and statistical. I t has the potential to justify the acceptance of many more minority indi viduals into select colleges based on their test performance, along with other important factors such as high school grade point average. Stephen Jay Gould (1995) reminds us that a test can be biased in at least two ways, culturally or statistically. Gould's distinction crystallizes several ideas regarding test bias. He explains that a standardized test may be culturally biased when one group (typically a minority populatio n) performs consistently lower than some reference population - typicall y, the White population. He adds that a test is considered statistically biased if two individuals (eg, one African American, one White) who g et the same test score nevertheless perform differently on some criterio n external to the test, such as school grades. One can extend Gould's argument to say that a test is culturally biased i f individuals from different ethnic groups interpret critical terms in m any of the test items differently. The consequence of this interpretive difference would be the observed mean differences in test performance. B uilding as well on Gould's statistical bias definition, there is at leas t one other sense in which a test can be biased. For example, if two ind ividuals get the same verbal test score, it is reasonable to assume that they should perform approximately equally well on all aspects of the ve rbal test itself. However, if two individuals with the same overall scor e - or, more generally, two ethnic groups matched on some total test sco re - should differ substantially on different subparts of a test, we wou ld say that the test is also statistically biased. Based on the results presented below, I assert that the SAT, as currently administered and analyzed, is both culturally and statistically biased in the ways described above. I also show how both cultural and statistic al bias can be partially ameliorated by scoring one half of the SAT, the hard part. The hard items are those that are often dependent on rare vo cabulary, whereas the easy items are dependent on terms that are typical ly more common. Background Studies of Cultural and Statistical Bias in the SAT Cognitive psychologists Freedle and Kostin (1987, 1988, 1990, 1997) use a measure called the standardized Differential Item Functioning (DIF) sta tistic to study ethnic bias in any standardized test. The respective proportions of White and African Americans who correctly answered that first item are computed, and then the difference in proportions is determined. This difference is then weighted by the n umber of African Americans scoring 200. Notice that both Black and White students are said to be of matched ability here because both groups rec eived the same score of 200. Next, for the same item, the same weighted computation is performed for a ll White and African American candidates who scored a 210. These steps (from 200 to 800) toget her yield sixty-one weighted computations, all applying to the first ite m One sums these sixty-one computations and determines their average va lue. This average is called the DIF score for the first item. All subseq uent verbal items are then examined by the same procedure, and a DIF sco re is assigned to each of them. A positive DIF score for an item indicat es that the African American population performs differentially better t han their matched-ability White peers. For example, White students may get 84 percent correct on some ea sy items, while African Americans get a slightly lower number, say 82 pe rcent, correct for that same item. Conversely, for some particular hard items, White students might get 30 percent correct whereas African Ameri cans might get a slightly higher score, say 31 percent correct. What is unusual about these effects is their highly patterned nature; that is, m any easy items show a small but persistent effect of African Americans' underperformance, while many hard items show their overperformance. Later I will make it clear that these small single-item bias effects beco me magnified, partly because the traditional scoring of a paper-and-penc il SAT gives equal weight to every item. In other words, a correct easy item carries the same weight as a hard item. I examine this assumption o f equal weight below in terms of its effects on ethnic bias. The Lexical Ambiguity and Cultural Familiarity Hypothesis The largest positive and negative DIF values occurred among analogies and antonyms, the verbal item types with the least verbal context. In contr ast, the smallest DIF values tended to occur for the reading comprehensi on items, which had the maximum verbal context. Freedle and Kostin (1990 ) observed two major effects: 1) DIF item values, both positive and nega tive, increase as the amount of verbal context decreases; and 2) within each item type (except for the reading items) the easy items typically r eceive negative DIF values, while the hard items typically receive posit ive DIF values. In short, test bias against minorities occurs primarily for easy analogy and easy antonym items. In this case, the fact that the correlation is positive indicat es that hard items are associated with positive DIF values, while easy i tems are associated with negative DIF values. Replication of the Ethnic Bias Pattern: A Brief Review Freedle and Kostin (1997) reviewed other background studies that replicat ed and extended the general pattern of DIF that they reported earlier. They also found, like Freedle and Ko stin, that analogies and antonyms yielded the largest DIF scores for eac h ethnic group, and reading items yielded the least bias. Schmitt, Doran s, Crone, and Maneckshana (1991) also reported similar verbal DIF findin gs for all three ethnic groups. Finally, Freedle and Kostin (1997) reana lyzed data reported by Raju, Drasgow, and Slinde (1993) regarding a 45-i tem vocabulary test for grades ten and twelve. Participants included 245 African American and 436 White students. Again, they found significant effects favoring the African American students for the hard vocabulary i tems and disfavoring the African Americans for the easy items. It is clear then that other researchers have replicated this ethnically b ased response pattern, with easy items generally being better performed on by the White majority and hard items generally being better performed on by each of the minority groups. Wh... |
csua.org/u/bwm -> www.jacksonville.com/tu-online/stories/030701/opl_5572812.html Opinion Buttons Wednesday, March 7, 2001 Story last updated at 4:41 pm on Tuesday, March 6, 2001 PUBLIC SCHOOLS: Teaching to test narrows the curriculum It's the time of year when schools across Florida administer the Florida Comprehensive Assessment Test. I have been teaching mathematics in the same Florida high school for almo st 20 years. I would like to share the position of the National Council of Teachers of Mathematics on high-stakes testing. The NCTM defines high -stakes tests as "tests that are used to make significant educational de cisions about children, teachers, schools or school districts." The NCTM January/February 2001 news bulletin states: "To use a single obj ective test in the determination of such things as graduation, course cr edit, grade placement, promotion to the next grade or placement in speci al groups is a serious misuse of such tests." The NCTM further states: "Far-reaching and critical educational decisions should be made only on the basis of multiple measures." Measuring students' abilities in a narrow range of mathematical and commu nications skills does not necessarily provide a true measure of a school . Using the FCAT as the major tool for grading and funding schools seemi ngly ignores everything else. It does, however, encourage schools to tea ch to the test. The NCTM warns that "teaching to the test narrows the curriculum." Less t ime spent on algebra and more time spent on FCAT skills may produce some one who can pass the FCAT but who probably now is a weak algebra student , who may then become an even weaker geometry student and so on down the line. Colleges and universities across Florida now have a disparate num ber of entering students taking remedial math. The FCAT has demoralized hardworking teachers, stressed students and is c reating hardship for the schools that need the most help. If you want to grade schools, th en measure the progress of their students across the curricula over the years. Teachers can only teach students who attend class regularly, are ready to learn and, most importantly, are willing to work at it. By the way, I'd like to see how w ell our state legislators would do on the FCAT. |