Table of Contents
- Cautions on the Use of Aggregate SAT Scores
- Family Income
- Academic Preparation
- North Carolina and the University of North Carolina System
- North Carolina School Systems and Schools
- Background on Recentering the SAT I Score
- Sources of Data for the Report
- Data Appendices (Adobe Acrobat Files):
Cautions on the Use of Aggregate SAT Scores *
As measures of developed verbal and mathematical abilities important for success in college, SAT scores are useful in making decisions about individual students and in assessing the academic preparation of individual students. Using these scores in aggregate form as a single measure to rank or rate teachers, educational institutions, districts, or states is invalid because it does not include all students. And in being incomplete, this use is inherently unfair.
For example, in order for one to make useful comparisons between states, of students' performance, a common test given to all students would be required. Because the percentage of SAT-takers varies widely among the states, and because the test-takers are self-selected, the SAT is inappropriate for this purpose.
The most significant factor in interpreting SAT scores is the proportion of eligible students taking the exam the participation rate. In general, the higher the percentage of students taking the test, the lower will be the average scores.
In some states, for example, a very small percentage of the college-bound seniors take the SAT. Typically, these students have b academic backgrounds and are applicants to the nation's most selective colleges and scholarship programs. Therefore, it is to be expected that the SAT verbal and mathematical averages reported for these states will be higher than is the national average. In states where a greater proportion of students with a wide range of academic backgrounds take the SAT, and where most colleges in the state require the test for admission, the scores are closer to the national average.
In looking at average SAT scores, the user must understand the context in which the particular test scores were earned. Other factors variously related to performance on the SAT include academic courses studied in high school, family background, and education of parents. These factors and others of a less tangible nature could very well have a significant influence on average scores. That is not to say, however, that scores cannot be used properly as one indicator of educational quality. Average scores analyzed from a number of years can reveal trends in the academic preparation of students who take the test and can provide individual states and schools with a means of self-evaluation and self-comparison.
By studying other indicators such as retention/attrition rates, graduation rates, the number of courses taken in academic subjects, or scores on other standardized tests one can evaluate the general direction in which education in a particular jurisdiction is headed. A careful examination of other conditions impinging on the educational enterprise, such as pupil/teacher ratios, teacher credentials, expenditures per student, and minority enrollment, is also important.
Summaries of scores and other information by state, college, or school district can be used in curriculum development, faculty staffing, student recruitment, financial aid assessment, planning for physical facilities, and student services such as guidance and placement. Aggregate data can also be useful to state, regional, and national education policymakers, especially in tracking changes during a period of time.
Scholastic Assessment Test (SAT) scores measure developed verbal and mathematical abilities necessary for success in college. Toward this end, SAT scores are useful in assessing the academic preparation of individual students and in making decisions about individual students. Using SAT scores in aggregate form as a single measure to rank or rate states, educational institutions, school systems, schools, or teachers is invalid because not all students take the SAT and those who do are self-selected. Comparisons of this kind are incomplete which makes their use inherently unfair. Consequently, rankings or residual rankings are not used in this report in compliance with The College Board and with professional standards for educational and psychological testing.
Aggregate scores can, however, indicate the preparation of groups of students who aspire to attend college. In addition, average scores analyzed for a number of years can reveal trends in the academic preparation of students who take the SAT. Consequently, this report includes the SAT performance of North Carolina students who took the test in 1998, as well as historical data on the SAT performance of North Carolina's students.
The data in this report are SAT results for students scheduled to graduate in 1998 and are these students' most recent scores regardless of when they last took the test. The results for North Carolina and the United States in this report represent the performance of public and non-public school students. North Carolina's results include performance of students in public schools, charter schools, North Carolina School of the Arts, and North Carolina School of Science and Mathematics.
This year North Carolina students made a mean gain of four points while the nation's mean SAT score improved by one point (see Figure 1). The North Carolina mean (average) total SAT score for 1998 college-bound seniors is 982. North Carolina students continue to show improvement each year, however, they are 35 points below the national mean. The gap of 35 points between the North Carolina mean and the United States mean is the smallest in 27 years. The gap between North Carolina and the nation has decreased by 48 points since 1972 and has decreased by 23 points since 1988.
North Carolina continues to close the gap on the Southeast as well. Compared to North Carolina's four-point gain, the Southeast mean total SAT score increased by two points in 1998. North Carolina is four points below the Southeast mean of 986.
Students in North Carolina public schools are faring even better when compared to public school students in the nation. For public school students in the nation the mean total SAT score is 1011 compared to North Carolina public schools' 981 (a difference of 30 points).
North Carolina students are closer to the nation on the verbal portion of the SAT than on the mathematics portion. North Carolina students' mean score on the verbal portion of the SAT was 490 compared to 505 for the United States (a difference of 15 points). On the mathematics portion of the SAT, North Carolina students' mean score was 492 compared to 512 for the United States as a whole (a difference of 20 points).
In North Carolina and the nation, males historically have higher mean SAT scores than females (see Figure 2). North Carolina males' mean total SAT score is 1002 compared to females' 967. The gap between North Carolina males and females has been 32-33 points the past four years, but increased to 35 points this year. The difference between North Carolina males and females is largely in mathematics where North Carolina males' mean score of 509 is 30 points higher than North Carolina females. North Carolina females only trail five points behind North Carolina males on the verbal scale.
North Carolina males trail males nationally by 38 points, while North Carolina females fall behind females nationally by 31 points. North Carolina males and females had a net gain on the mean total SAT of three points compared to their counterparts nationally. Nationally, males' mean total SAT improved by three points while North Carolina males improved by six points; females nationally improved by one point while North Carolina females improved by four points.
Whites and Asians in North Carolina schools and in the nation typically score higher than other ethnic/racial groups (see Figure 3). Of all the ethnic/racial groups represented in North Carolina, only Hispanic students' score higher than their national counterparts. North Carolina Hispanics' mean total SAT score was 984 compared to 916 for Hispanics nationally (a 68 point difference). Hispanics of all backgrounds (Mexican, Puerto Rican, or Latin American) in North Carolina had higher mean scores than their national counterparts. Hispanics, however, make up a very small proportion of the total SAT test takers in North Carolina, representing only one percent compared to eight percent nationally.
All other ethnic groups, American Indians, Asian Americans, Blacks, and Whites in North Carolina trail their national counterparts in SAT performance. Of these groups, Blacks in North Carolina are the closest to their counterparts nationally (a difference of 23 points). In North Carolina the mean total SAT score for Blacks was 839, a five point increase over last year. This compares to Blacks nationally who increased three points to 860. Black students have demonstrated improved performance in North Carolina, while representing a much higher proportion of SAT test takers than in the nation (20% in North Carolina versus 11% nationally).
North Carolina's American Indian students, although improving performance, are the farthest behind their national counterparts (a difference of 57 points).
In North Carolina, as in the nation, the higher the family income the higher the student's mean total SAT score (see Figure 4). There is very little change from year to year in the mean within each family income category. The relative difference in mean total SAT score between family income categories is also fairly stable from year to year.
The more academic credits students have in six subject areas (Arts and Music, English, Foreign and Classical Languages, Mathematics, Natural Sciences, Social Sciences and History), the higher their mean SAT scores (see Figure 5). Students in North Carolina are earning one-half additional academic credit on average compared to students in 1994. While the mean SAT scores of students in the nation in each range of earned academic credits have fluctuated over the last five years, the mean scores of North Carolina students in two of the higher levels of academic credits earned (18-18.5 and 19-19.5) have steadily decreased over the last five years (see Figure 5). This trend may be a result of a more diversely prepared group of students taking the SAT which is reflected in the percent of SAT test takers with 20 or more academic credits increasing from 39% in 1997 to 47% in 1998.
Typically, the higher a student's high school grade point average (GPA), the higher the student's mean total SAT score (see Figure 6). North Carolina students with a high school GPA of A-, A, or A+ are further behind their national counterparts than students with a B or C average. North Carolina students with a high school GPA of A-, A, or A+ trail their peers nationally by 65, 60, and 51 points respectively. Additionally, North Carolina students with a GPA of A-, A, or A+ represent 43 percent of students compared to 38 percent nationally. North Carolina students with a GPA of B are 41 points behind their peers nationally and represent 44 percent of students compared to 48 percent nationally. Students in North Carolina with a GPA of C are only 28 points behind their peers nationally and represent 13 percent of students in North Carolina and in the nation. Examining the information on student GPA and mean SAT scores, it appears that North Carolina teachers' standards for A's and B's are lower than their peers nationally, while their standards for C's is comparable to their peers nationally. A small decline in the mean SAT scores for students with a GPA of A-, A, or A+ and a small increase in the mean SAT scores of students with a GPA of C over the last five years (see Figure 6), further implies this discrepancy in standards.
There is a b, positive relationship between the average performance of students on the North Carolina end-of-course tests in a high school and the mean total SAT score for that school (see Figure 7). The Pearson correlation between the performance composite and mean total SAT score by high school was 0.88 on a scale of 1.0 to +1.0. This relationship was determined by plotting a high school's performance composite against its mean total SAT score. The performance composite is the weighted average of the percent of students at or above level III on end-of-course tests (i.e., students mastering the course content). The performance composite is based on student performance on six end-of-course tests: Algebra I; Biology; Economic, Legal, and Political Systems (ELPS); English I; English II; and U.S. History.
The mean total SAT score of North Carolina students graduating in 1997 was 978, while the University of North Carolina system mean for first-year students was 1060 (1997 is the most current year for which comparable data are available). The fact that students entering the University of North Carolina system have a higher mean total SAT score than students graduating high school is expected since many students who do not perform well on the SAT choose other post-secondary options including community college and full-time employment. University of North Carolina institutions, however, serve a wide-variety of student abilities as evidenced by the institution averages which range from 825 to 1220 (The University of North Carolina, 1998).
Most people assume there is a negative association between the percent of students taking the SAT and the mean SAT score. This association is true when the percent of students taking the SAT and the mean total SAT scores for states are compared (see Figure 8). However, the opposite association occurs when the percent of students taking the SAT and the mean total SAT score for public school systems and public schools are correlated (see Figures 9 and 10). The Pearson correlation between the percent of students taking the SAT and the mean total SAT score is 0.39 for public school systems in North Carolina and similarly the correlation is 0.41 for North Carolina public schools. These results mean that schools and school systems in North Carolina cannot assume that their scores were better or worse because the percent of students taking the SAT changed. In fact, 51% of all schools and school systems in the nation had a change in their mean verbal or math SAT of plus or minus 10 points (The College Board, 1998). This fluctuation in mean SAT scores means that school systems and schools should take into account other factors such as course-taking patterns, content of the curriculum, and course standards when attempting to explain changes in mean SAT scores.
Background on Recentering the SAT I Scores
The College Board recentered the score scale of the SAT I, re-establishing the original mean score of 500 on the 200-800 scale in order to maintain the SAT's statistical integrity and predictive validity. The scale had not been recalibrated since 1941 when it reflected the norm of some 10,000 students from predominantly private secondary schools who applied to the nation's most selective private colleges and universities. As mean scores shifted below 500, the score distribution became stretched in the upper half and compressed in the lower half.
Now that scores are recentered on the renormed SAT I, they reflect the more than two million students who take the test today. They also reflect a more diverse college-bound population than the group who took the SAT in 1941.
Although a student's score may change after recentering, the rank order of individual scores, expressed as percentiles, remains the same. What is more, a specific score on the verbal test now has the same relative position and meaning as the same score on the math test. For example, a 450 on verbal and math signifies comparable performance in both areas. Before recentering, a score of 450 represented above-average performance on verbal and below-average performance on math. While recentering permits legitimate comparisons of verbal and math scores and reduces earlier confusion, it has no effect on historical score trends, or on the difficulty level of the test and the relative standing of students to each other.
Sources of Data for the Report
The data in this report are from three primary sources: (1) National Report 1998 College-Bound Seniors: A Profile of SAT Program Test Takers and profiles from earlier years (The College Board); (2) North Carolina Report 1998 College-Bound Seniors: A Profile of SAT Program Test Takers and profiles from earlier years (The College Board); and (3) a data file of individual student scores for the state's 117 public school systems, charter schools, North Carolina School of the Arts, and North Carolina School of Science and Mathematics. The data file was prepared by Educational Testing Service in cooperation with The College Board. SAT scores are reported each year for students scheduled to graduate. Only the most recent scores of these students are reported, regardless of when they last took the test.
The University of North Carolina. (1998, April). Statistical abstract of higher education in North Carolina, 1997-98 (Research Report 1-98). Chapel Hill, NC: Author.
The College Board. (1998, September). College-bound students set records in racial and ethnic diversity, precollege credit, and grades, but College Board sees growing disparities among subgroups. New York: Author.
Report Produced By:
Public Schools of North Carolina
State Board of Education
Department of Public Instruction
Instructional and Accountability Services
Division of Accountability Services