Failed education fads seemingly never die. They may fade away for a few years but true believers usually revive them, sometimes under a new name.  So it is with portfolio assessment, which was briefly the rage in the 1990s when education progressives heralded it as an “authentic” way of testing student progress.

Now, portfolios are re-surfacing as a proposed alternative to standardized testing mandated by No Child Left Behind (NCLB) just as some independent researchers are finding that, after its first five years, the federally mandated testing may be starting to spur modest gains in student achievement.

Portfolios are collections of any sort of student work done over time – essays, book reviews, drawings, laboratory reports, research projects in any subject.  Few education experts doubt the value of teachers examining student portfolios in individual classrooms. However, many doubt the validity of portfolios as a measure of schoolwide, district, and state educational progress.

The problem with portfolio assessment for accountability purposes is its subjectivity. Essentially, it makes schools accountable only to themselves. No one need bother with external oversight using objective yardsticks.

The National Education Association, which opposes any form of merit pay or evaluation of teachers based on their students’ academic growth, now seeks to pressure states to use portfolio assessment to replace or minimize the standardized testing that has been used to hold schools accountable for improving math and reading results in grades 3-8.

But, as the Education Consumers Foundation’s J.E. Stone notes, “The two states that attempted to build their accountability systems around portfolio assessments – Kentucky and Vermont – had to abandon the project. In both cases, the systems virtually collapsed of their own weight.”

In Kentucky, a 1995 legislatively commissioned study of the portfolio assessment required by the state’s 1990 education reform act revealed huge problems. Among them were lack of controls to ensure reliability, and the great variation in the assistance students received from teachers, peers, and parents in performing their “authentic” tasks.

Kentucky subsequently reinstalled standardized, norm-referenced testing.

About the same time, a RAND Corporation team reached similar conclusions after looking at portfolio assessment in Vermont. One school or teacher might require one kind of project; another school or teacher an entirely different sort. Portfolios were time-consuming, expensive, and robbed teachers of time to teach basic skills.

In his 2005 book, “Kill the Messenger: The War on Standardized Testing,” Richard Phelps pointed out that portfolios are “open invitations to cheating.” He continued:

“If a student turns in a portfolio for a statewide, standardized assessment, who is to know if the essay enclosed is written by that student, or her mother, or by someone in New Zealand who posted it on the Internet?”

It is true that under the current NCLB testing regimen, some school districts and states have been able to manipulate scores to make their students’ performance look better than it is. Where large NCLB gains stack up against flat or falling scores on the National Assessment for Educational Progress, suspicions may be justified.

Still, reverting to portfolio assessment would make accountability not just hard to achieve, but virtually impossible.