fenlo012: November 2011 Archives

The use of ACT and SAT scores in the college admissions process has always been debated. Considering the correlation between admissions tests and college grades are typically below .5 as the Lillenfield text points out, I can see why the use of such scores is so criticized. Many students, like Danielle Rettinger who was interviewed in an NPR report (see link), feel that the tests do not adequately show academic performance. The College Board argues that standardized tests are necessary for colleges to use considering the number of applications they review. Those who argue against the use of SAT and SAT scores say that the number is not an adequate predictor of a student's ability and that the test is particularly biased against women and minorities. I see a problem in the reliability of standardized testing. For example, those who do not do well on the SAT usually score a better equivalent score on the ACT. The SAT and ACT also vary on what they test. Although they test in some similar subjects, the SAT focuses more on grammar than the ACT whereas the ACT tests harder math skills than the SAT. Some colleges have chosen to do away with the ACT and SAT requirement. Bates College in Maine has found that the difference in graduation rate between students who submit scores and those who do not is less than .1 percent indicating that the tests may not be the best predictor of success. I do not think that SAT and ACT scores should be the sole predictor of one's intelligence. As Howard Gardner illustrated, intelligence can come in many forms. Some students are not successful at test taking but might show high intelligence in a different area of academics. Although it would be more difficult during the admissions process, I believe that ultimately, colleges could find more value in looking at overall intelligence rather than a number scored on a standardized test.

NPR article about ACT and SAT scores

I have always struggled with the idea that a single number can determine the intelligence of an individual. After reading chapter 9 and watching the following video, I realized that IQ should not be used as a single marker of one's overall intelligence.

Although IQ scores are valid, there are aspects of IQ testing that must be looked at closely. IQ can predict some aspects of human intelligence. For example, good health literacy and success at mentally demanding occupations both have a positive correlation with IQ. And IQ tests are also valid because the correspond with other modern tests.
However, when considering the reliability of an IQ test, we must be very careful. In general, IQ scores remain stable throughout adulthood. But IQ scores are not set and can drastically change over time. So as the man in the video states, it is unfair to set the standards for a student's education based on his standardized test scores from one point in time, in this example, fourth grade. In order to make the testing more reliable, the tests must follow test-retest reliability. In this instance, a student's future could be determined by his scores on one IQ test, but it would be more fair for scores to be taken many times in order to insure reliability. I have personally experiences the difficulties one time IQ tests can cause for an individual. For high school I was asked to take an entrance exam which would ultimately place me in the "correct" math and science classes. I was placed in the middle level math class when in reality, I could have taken the upper level math class. Because my placement was based solely on one test, it became an unreliable marker of my ability. In this situation, I was unfortunately set back because I didn't score well enough on one test. While IQ is helpful and usually a good marker of intelligence, it is important to remember some of the basic scientific thinking principles. Tests must be replicated to ensure reliability and to limit the number of confounding variables, such as specific difficulties on test days due to illness or personal situations.

About this Archive

This page is an archive of recent entries written by fenlo012 in November 2011.

fenlo012: October 2011 is the previous archive.

fenlo012: December 2011 is the next archive.

Find recent content on the main index or look in the archives to find all content.