Facts of the Matter


Sunday, December 2, 2001

Assumptions about
science education
fall short

On a flight back to Hawaii from Salt Lake City recently I had a conversation with the parents of two teenagers. They complained of low test scores and placed the blame on their state's low expenditures for education. I said that Hawaii probably spends more per student but had comparable test scores.

It seems like a no-brainer that more money means better education, but it's all too easy to jump to conclusions from limited data. Besides, common sense has been known to lead us astray so I dug a little deeper into the test data to see if they corresponded with common sense.

The tests I'm referring were administered by the National Assessment of Educational Progress , also known as "the Nation's Report Card." It is the only nationally representative and continuing assessment of what America's students know and can do in eight subject areas including science. They were administered to grades four, eight, and 12.

Nearly a quarter of a million students nationwide took the tests last year. At the same time students and teachers were polled about their instructional experiences and the school environment. The test questions came from earth, physical and life sciences and were based on the three characteristic elements of science: conceptual understanding, scientific investigation and practical reasoning. The questions were a combination of multiple-choice and constructed-response with half of the students also performing a hands-on task.

The percentage of Hawaii's students in grades four and eight who scored at proficient levels in science is about half of the national average. About one in seven in both grades scored at proficient levels, and only one in a hundred scored at advanced levels. Sixty percent of Hawaii's eighth graders and 49 percent of fourth graders scored below the "basic" level.

By comparison with Utah, the test scores of Hawaii's eighth-graders were near the bottom, tied at number 40 out of 41. Utah was 14th from the top. Common sense says such factors as expenditures, teacher salary and class size affect student learning, yet the numbers for Hawaii and Utah indicate otherwise.

Utah's expenditures per student were only two-thirds that of Hawaii's, yet Utah's test scores were much higher. Hawaii ranked 28th out of 54 states and jurisdictions, Utah was 53rd.

Teacher salaries are lower in Utah, but assuming a 22 percent cost of living allowance the two states are identical, but Utah's test scores were much higher. Hawaii's teachers earn the 18th highest salaries, Utah is 42nd.

Class sizes are 25 percent bigger in Utah, and Utah's test scores were much higher. Nationally Hawaii is 12th highest, Utah is second.

These results seem to fly in the face of common sense, but there are other things to consider, namely statistics. But even that can be confusing.

Upon statistical analysis, the nationwide data show very little correlation between the test scores and class size or education dollars. Greater expenditures, higher teacher salaries and smaller class sizes do not appear to be related to student science scores.

Statistical correlation is easily misconstrued and the results misused. Correlation does not mean that there is a cause and effect relationship. It measures the amount of scatter, and there is much scatter in these data. Despite a low correlation, there is a trend in all three cases in the direction we would expect. Higher test scores do correspond with higher salaries, greater expenditures and smaller classes, but with much scatter. This means that common sense is correct, but there is no guarantee that smaller class sizes, greater expenditures, or higher salaries will lead to higher test scores. It means that money alone does not buy test scores and we need to look for solutions to the education problems elsewhere.

A noteworthy point is that this study only included public school students. Since Hawaii ranks among the highest in the country in proportion of private to public school students it would be interesting to see the data include nonpublic schools if we are measuring overall student proficiency.

The results of the concurrent survey on educational experiences and environment are interesting and should provoke conversation among students, educators, and parents. Here's a summary:

>> Teachers' undergraduate major is related to student science achievement at grade eight.

>> Certain types of computer use in the classroom are related to science achievement.

>> Students scored higher when teachers had them work together on science activities at least weekly.

>> Science courses taken are related to achievement at grades eight and 12.

>> Certain types of computer use had a positive relation to performance at grade 12.

>> Television/video watching is associated with student performance at grades four, eight and 12.

>> Internet use in the home had a positive relation to student performance at grades four, eight and 12.

For more information see a summary with graphs on my Web site at or visit the NAEP Web site at

We could all be a little smarter, no? Richard Brill picks up
where your high school science teacher left off. He is a professor of science
at Honolulu Community College, where he teaches earth and physical
science and investigates life and the universe.
He can be contacted by e-mail at

E-mail to Business Editor

Text Site Directory:
[News] [Business] [Features] [Sports] [Editorial] [Do It Electric!]
[Classified Ads] [Search] [Subscribe] [Info] [Letter to Editor]

© 2001 Honolulu Star-Bulletin