Top students surpass nation's top numbers

There has been considerable criticism that the top students graduating from Indiana high schools do not measure up with their counterparts from other states. It is important that we do more to improve student performance, but first we must agree on how we are going to measure performance, then we will know whether we are seeing improvement or not. Furthermore, it is important than we expect all students to improve not just some. If students came to school with the same ability, parental support, motivation, and the other external factors which affect how they perform, then it would be fair to compare students as if they were the same and the only factor that made a difference was the school in which they were enrolled. However, seldom are all of those factors equal.

The recent release of Scholastic Aptitude Test (SAT) results is an excellent example of these unfair comparisons. In Indiana, there is always a question about the large number of students taking the SAT. Is it fair to compare Indiana with 60% of the students taking the SAT in 2000 with North Dakota who has only 4% of the students take the SAT but has the highest scores on both the verbal and math test?

We made a comparison using a smaller number of Indiana students to show the difference. From Noblesville High School students, we combined Verbal and Math SAT scores instead of using the individual ones because it is easier to show comparisons. Instead of using four percent of our student scores, we used the SAT scores from the top ten percent of the 2000 graduating class. These results are shown in the table below along with the results from the top 1, 10, 20, 25, 30, and 40 states.

Top	Percent	Combined Score
1	4%	1,197
10	7%	1,164
20	8%	1,138
25	13%	1,125
30	18%	1,111
40	30%	1,087
National	44%	1,019
NHS	10%	1,257
NHS	79%	1,005

Source: The College Board at http://www.collegeboard.org/press/senior00/html/table3.html

As you can see from the table, the top ten percent of the students at Noblesville High School scored better than all the other states. Would the critics of Indiana student performance accept this data? We suspect that they would say it is not a proper use of the data. We would agree with them, but we would state that this is not any less appropriate than how they are using the data.

When you add the other 69% of the students at Noblesville High School who took the SAT, the results are quite different. We had been very pleased with our SAT results over the last five years because we had a steady increase in SAT scores. We had attributed the increase to the change in academic offerings at our high school and the increase in the number of students taking those courses. In 2000 we had a decline from the previous year, we decided to analyze our results differently. We want to share those results in terms of an improvement model.

What SAT results should we be considering?

We decided that the only way that we could determine whether student performance on the SAT was improving was to compare results for the same students from one year to the next. In reviewing our records, we found that we had scores on both the Preliminary Scholastic Aptitude Test (PSAT) and the SAT. For the class of 2000, we had 317 take the PSAT as juniors (82%) versus the 291 (79%) take the SAT as seniors. Although these two groups did not have the same students, there was such a high percentage of students within both testing groups that it was fair to compare their average results. Shown in the table below are the PSAT and SAT results by graduating class for the last six years.

	PSAT		SAT
Class	% Tested	Combined	% Tested	Combined	Change	z Score
2000	82%	937	79%	1,005	68	1.64
1999	60%	1,012	79%	1,029	17	-1.51
1998	61%	994	75%	1,033	39	-0.15
1997	53%	988	68%	1,029	41	-0.03
1996	67%	965	70%	1,007	42	0.03
1995	47%	881	72%	923	42	0.03
				Mean	41.5
			Standard Deviation		16.2

We converted the PSAT score to a score similar to the SAT by dropping the decimal point on the PSAT. For example a 51.9 became at score of 519, this is a practice that school counselors have been using for years. In looking at the growth in student performance from one year to the next, we find there was an improvement. By computing average score and standard deviation for the six years, we are determining what differences are most significant. Now our findings give new insight. The year that we had the highest scores was also the year that we had the least amount of growth (-1.51). This last year has been the year with the greatest growth (+1.64).

According to The College Board 2000, on the average juniors taking the PSAT in October had a gain of 10 points higher in verbal and 12 points higher in math (http://www.collegeboard.org/sat/cbssenior/html/stat00e.html). The 68-point gain by the Class of 2000 was more than three times the national gain as compared to 17-point gain of the Class 1999. Being able to understand and use this information is what we need to be able to do if we are ever going to set expectations that will improve student performance.


How should we use the test results?

There are number of things that we can do to use these results to improve student performance. Some of these are:

• expect performance to improve from taking the PSAT one year to taking the SAT the other year;
• get a higher percentage of students to take both testings (the Class of 2000 could be the most accurate assessment of student performance);
• provide interventions on an individual basis to guarantee the maximum student improvement; and,
• look at student improvement rather than comparing and ranking.