USA Today education reporter Greg Toppo recently hit the reportorial nail on the head with his take on education research. After attending the oh-so-enlightening American Educational Research Association's (AERA) annual conference this month, Toppo rightfully questioned the usefulness of much of the research that was presented.

Toppo snagged skeptical quotes from Rick Hess and Grover Whitehurst, but of course, balanced it all out with an interview with AERA President William Tate, who vigorously defended the status quo. Tate asserted that a dubious clunker from Teachers College on reducing graduation rates "should be in the hands 'of every superintendent in America.'"

Hess, who has taken to hanging out at the hotel bar during the AERA convention rather than subjecting himself to any actual sessions, offers a good rundown of some typically politicized and/or navel-gazing education research from AERA's 2007 annual conference here. By only looking at one day's worth of papers, though, Hess failed to mention important work like "Teaching Graduate Students to Become Effective Proteges." In most academic fields, this paper's topic would be the subject of professors' water-cooler talk while taking a break from doing their real research--but in education it is the research! Or, take for example the forward-thinking symposium on "21st-century literacy," including "Defining Online Reading Comprehension: Using Think Aloud Verbal Protocols to Refine a Preliminary Model of Internet Reading Comprehension Process." For the ultimate in trendy research, however, we have no choice but to nominate "Is That a Wedding Ring? A Look at the Panopticons of Identity Politics Lived by Closeted Gay School Administrators Serving Homophobic Communities."

A few notches above the standard AERA fare stands a paper delivered by the ever-insightful education economist, Dan Goldhaber. He provides fresh evidence on the relationship between states' licensing tests and teacher effectiveness. For reasons that don't speak well of either states or their testing contractors, neither party has shown much interest in learning if licensing tests and the highly variant, minimum passing scores (set by individual states) serve as an appropriate obstacle into the profession. Goldhaber mines a North Carolina value-added dataset of elementary teachers from a 10-year period. What he brings new to this issue is controlling for the nonrandom assignment of teachers to classrooms. Researchers have never been quite sure if teacher test scores do matter--as less sophisticated studies have shown--or if the differences are simply reflecting the fact that higher scoring teachers tend to be assigned to higher performing classrooms.

In accounting for "non-randomness", Goldhaber finds that, indeed, teacher tests do predict teacher effectiveness. Test score differences are about as important as the relatively powerful measure of teacher experience, those differences found between a novice teacher's effectiveness compared with the effectiveness of a teacher with a year or two of experience.

There are some twists on his findings, though. Goldhaber finds that another state (Connecticut) with higher minimum passing scores than North Carolina, doesn't gain any particular advantage from its relatively higher scores. In fact, the state is likely turning away teachers who would have been quite effective (termed "false negatives"). This finding will be music to the ears of states that are constantly reminded by groups such as ours that their passing scores are embarrassingly low.

Further, not all tests are created equal, with some proving to be better predictors of performance than others. One test, entitled the Praxis II Curriculum test, correlated quite nicely with teacher performance. That test requires teachers to solve instructional problems set in the context of the four primary subject areas encountered in the elementary classroom. The other test, the Praxis II Content Area Exercises, , showed little to no correlation. That test requires teachers to discuss an instructional approach, develop an instructional goal or solve an instructional problem.

Now a sidebar note from research contained in our soon-to-be-released State Teacher Policy Yearbook. There's a widely held perception that states put a lot of stock in their licensing tests. While tests are most certainly more important now than they were 30 years ago, it's nonetheless an area in which states are willing to show their largesse. The same states that would never consider granting a waiver for a course such as Foundations in Education will offer a host of regulatory loopholes exempting teachers from having to take their tests. For example, if a teacher has a major in the subject, many states have a policy that says that no subject area test is needed. What about a teacher who has a few years of experience? No need for a test. It's one more piece of evidence showing that states continue to vigilantly protect the interests of higher education, while showing half-hearted interest in applying more meaningful measures of teacher quality.