The Consequence of Banning High-Stakes Testing in (Science)

American education in general, and science education specifically have been radically and negatively impacted by high-stakes testing.

High-stakes testing, as set forth in the No Child Left Behind Act (NCLB), is the idea that the pressure of such tests will increase student achievement.  But one of the major studies cited here finds that the pressure created by high-stakes tests has no important influence on academic performance (Nichols, Glass & Berliner, 2005).

In this study, which was published in 2005, the researchers found that:

  • States with greater proportions of minority students exert greater pressure, and thus will disproportionately affect minority students.
  • High-stakes pressure is negatively associated with students in middle school moving on to the last year of high school.  The result is that many kids are kept back, or simply drop out of school.
  • Increased pressure did not result in gains on the National Assessment of Educational Progress (NAEP) reading test for elementary or middle school students.

These researchers conclude that

there is no convincing evidence that the pressure associated with high-stakes testing leads to any important benefits for students’ achievement. They call for a moratorium on policies that force the public education system to rely on high-stakes testing.

In another research study, Noddings reported that

although evaluation of student learning is required for accountability, high stakes testing is not required and may even be counterproductive. It also questions whether the goals of the ‘No Child Left Behind Act’ are reasonable and contends that, if they are not, there may be no justification for imposing punishments and sanctions on children and schools unable to meet them. Moreover, high stakes testing may be incompatible with many defensible aims – among them, critical thinking.

These two important ideas, that high-stakes testing has not been shown to improve student achievement, and that high-stakes testing is not required to assess student learning lead us to support the notion that

The consequence of eliminating high-stakes testing would be the immediate removal of the pressure on students and teachers to be controlled by a Federal policy that is deeply flawed, and based on a factory mentality of education that does not fit with 21st Century society.  The implementation of high-stakes testing has had more impact on what and how teachers teach than any curriculum innovation of the past 50 years.  Teachers are pressured to teach to the test, and in a case reported yesterday on this blog, Florida science teachers are throwing out hands-on and inquiry activities so that they can spend more time teaching to the end of year test (FCAT).

The consequence of eliminating high-stakes testing would mean that the likely hood that cheating on tests, by students and teachers would go by the wayside.  According to the Georgia Governor’s three-volume report, the Atlanta cheating scandal was caused by “a culture of fear, intimidation and retaliation that spread throughout the (Atlanta) district.”  That culture of fear was directly related to the pressure put on administrators, teachers, and students to make sure students scored high on the end-of-year tests at any costs.

The Atlanta scandal is not the first time cheating has happened.  In their book, Collateral Damage, Nichols, and Berliner report that cheating in American schools has occurred because it helps one gain advantages or avoid negative consequences.  In the present NCLB environment, schools can be sanctioned, or taken over, teachers fired, and students told they should go somewhere else to get a better education.

In the Atlanta case, cheating was widespread, and had gone on for several years, all during the enactment period of the NCLB Act.  So far, eleven educators’ teaching licenses have been revoked, and each faces criminal charges.  Another 180 teachers and administrators face sanctions and criminal charges.

I am not condoning the behavior of the educators in Atlanta, but I am using this case to reflect on the research that was done by Nichols, Glass and Berliner in 2005,  and their conclusion calling for a moratorium on policies that force the public education system to rely on high-stakes testing.

The consequence of eliminating high-stakes testing in science would enable science teachers to focus on science standards that encourage innovation, creativity and science-inquiry.  Teachers would also be free to relate science to the lived experiences of their students, to involve students in engaging scientific projects, and use the wealth of resources and curricula that have been developed by the leading science educational organizations in the Nation such as TERC and the Concord Consortium.

Science teaching, as well as other areas of the curriculum, would flourish with the banning of high-stakes tests.

But how would students and teachers be assessed if we didn’t use high-stakes tests.  That’s for tomorrow.

Resources:

High-Stakes Testing and Student Achievement: Problems for the No Child Left Behind Act, Sharon L. Nichols, Gene V. Glass and David C. Berliner, Arizona State University, Education Policy Studies Laboratory, 2005

High-Stakes Testing: Opportunities and Risks for Students of Color, English-Language Learners, and Students with Disabilities, Jay P. Heubert, Teachers College, Columbia University

Science Scores on International Assessments: The Sky is Falling

In yesterday’s post, I described Science Debate 2008, and efforts to engage the two major candidates for President to answer 14 important questions about science.  The one question that focused on science education was as follows:

A comparison of 15-year-olds in 30 wealthy nations found that average science scores among U.S. students ranked 17th, while average U.S. math scores ranked 24th.  What role do you think the federal government should play in preparing K-12 students for the science and technology driven 21st Century?

Each candidate discussed this question (you can read each candidates’ response by following this link and then scroll down to question #4).  You can decide for yourself which candidate provided the kind of answer that will improve science education in the nation’s schools.   To give you a flavor for what they said, here are two excerpts from their answers.  One of the candidates said that:

All American citizens need high quality STEM education that inspires them to know more about the world around them, engages them in exploring challenging questions, and involves them in high quality intellectual work. (please note: I added the link to the STEM education coalition)

The other candidate said that:

America’s ability to compete in the global market is dependent on the availability of a skilled workforce.  Less than 20 percent of our undergraduate students obtaining degrees in math or science, and the number of computer science majors have fallen by half over the last eight years.  America must address these trends in education and training if it hopes to compete successfully.  

Note that the question on science education was couched in the language of international assessments, e.g. “A comparison of 15-year-olds in 30 wealthy nations found that average science scores among U.S. students ranked 17th, while average U.S. math scores ranked 24th.”  What does this mean?  Is the sky falling?

No. Large-scale international assessments of student achievement receive a great deal of attention when the results become public.  Normally, the results comparing countries are reported in a fashion similar to standings in professional sports, where 1 is at the top, which is typically Singapore, followed by lower scoring countries, and as suggested in the question, placing the U.S.A. 17th out of 30.  

There is a real problem in using results to compare one country to another.  As some researchers have pointed out, the scores reported are averages for the country of the students who took the test.  Often the the differences between average scores from country to another are not significant, BUT politicians, educators and the public see that if their country is not NUMBER ONE, “the sky is falling.”

And its not just a concern expressed by U.S. politicians.  Svein Sjoberg of the University of Norway reports (in a study–Real Life Challenges: Mission Impossible) that results on the PISA test (Programme for International Student Assessment) of students in Norway provided “war-like headings” in most of Norway’s newspapers.  In fact the commissioner of education of Norway was quoted as saying, “Norway is a school loser, and now it is well documented.” 

So, when U.S. students score 17th on an international test, policy makers make the claim that science education in the U.S. is in free-fall, and needs to uplifted.  Remember, that the score used on these tests is an average.  There are more than 15,000 independent school systems in the U.S. and to use an average score on a science test (typically comprised of 40 – 60 questions) does not describe the qualities or inequalities inherent in the U.S.A.’s schools.  David Berliner (in a research study entitled Our Schools vs. Theirs: Averages That Hide The True Extremes)points out that the TIMSS (Trends in International Mathematics and Science Study) data for the U.S.A., when analyzed by socioeconomic levels, shows great disparities and inequalities.  For example, schools in the most affluent neighborhoods do well on these tests, but schools in poorer neighborhoods do not.  And Berliner points out that scores on international tests will not change unless the inequalities in the schools are fixed.

Results on these international achievement tests, taken out of context, might not be the best way to assess how well science is taught in any country’s schools.

How can school science be assessed that will help us close the gap between schools?  What suggestions do you have?  Do you think that these international test results are valid ways of assessing school science learning?