高校应用过程中中学生对人工智能的看法

Jeff Schiel Introduction Examples of college readiness indictors include applicants’ high school grades, essays, andcollege admissions test scores. Of these, high school grades are most prominent in the collegeadmissions process. Recently, the National Association for College Admission Counseling(NACAC) reported that high school grades were the most important factor in admissionsdecisions, with 74% of its member colleges viewing grades as considerably important. Over However, grades, despite their prominence, have imperfections. Evidence of a phenomenonknown as grade inflation, which occurs when there is an increase in grades over time but noconcomitant increase in ability (Bejar & Blew, 1981), has been around for many years. Forexample, ACT presented evidence of this phenomenon thirty years ago (Ziomek & Svec, 1995) Concerns over grade inflation continue, and recently the accuracy of high school grades as ameasure of student achievement has been questioned. For example, a study of grade inflationthat used longitudinal data from students in Washington state found significant changes overtime in grades and a noticeably weak relationship between grades in math and achievement,which was measured by scores on state-administered assessments (Goldhaber & Young,2024). In addition, the study found a significant increase in the proportion of students receivinggrades of A and B, even though these students were not meeting grade-level expectationsbased on their state assessment scores. The authors caution, however, that it is unclear how to Different types of grading systems occur in secondary education and include the traditional A–Fsystem, as well as proficiency grading, in which a student’s goal is to achieve mastery in each of a series of learning standards. Because many colleges and universities have chosen to makecollege admissions test scores optional, there are new challenges involved in the assessment of To summarize, grades appear to be gaining prominence in college admissions, even as theiraccuracy is being questioned and their interpretation is becoming more challenging. Moreover, itis possible that students’ use of artificial intelligence (AI) in their high school coursework hassome bearing on the accuracy of grades as a measure of student achievement. The use of AI iscommon among students and is increasing. For example, one study reported that in 2024, 26% Current Study College admissions decisions for high school students could be influenced by the interplayamong college readiness indicators and their perceived importance, the perception that highschool grades have declined in accuracy, and the influence of AI on grades and other •How do students view the importance of high school grades and other college readinessindicators? In this study, high school students shared their perspectives through an online survey on theimportance of college readiness indicators and whether they believe AI can influence aspects ofthe college application process. A total of 4,903 students responded to the survey. Additional Results Approximately two thirds (69%) of students in this study indicated that high school grades arevery important in the college application process (see Figure 1 and Table A2). Grades wereviewed as more important than other indicators, such as essays (53% indicated these are veryimportant), admissions test scores (51%), letters of recommendation (43%), and supplemental Mean importance ratings for each of the college readiness indicators are denoted by the whitesquares in Figure 1. Mean ratings ranged from 2.62 (high school grades) to 1.87 (supplemental Although high school grades were perceived as important, they were also perceived as beingable to be influenced by AI. Nearly three quarters (71%) of students at least slightly agreed thathigh school grades no longer accurately reflect student performance, because students can useAI to cheat (mean agreement rating = 4.20; Figure 2 and Table A3). In addition, 81% at least When considering the use of AI in the college application process, 70% of students at leastslightly agreed that AI provides an unfair advantage for some students (M= 4.15), and 69% atleast slightly agreed that students who use AI in this manner have a significant advantage over It is possible that students who are higher achieving academically would have differentperspectives on the importance of college readiness indicators and AI’s influence in the collegeapplication process than would students who are lower achieving academically. To investigate Students gave relatively high importance ratings to high school grades, irrespective of ACTscore category. However, students whose test scores were in the top quarter had differentviews of AI’s influence on high school grades: They were significantly less likely than their lowerscoring peers to agree that students who use AI in their high school coursework get better .0001,d=−0.21).1Higher scoring students were also signif