
Photo by Michelle Patterson, The Lexington Herald-Leader.
Education writers often try to measure   the success of their local   schools by analyzing such things   as test scores, class sizes and poverty   rates. But many reporters overlook one   of the most influential factors in   children’s school success: the quality   of teachers.
A primary reason why education   journalists tend to omit this measure   from their reporting is the difficulties   they confront in finding reliable data   about teacher qualifications.
But as we at The Lexington (Ky.)   Herald-Leader found, it can be done.   The Lexington Herald-Leader published   a four-part series last November   that demonstrated how our state’s nationally   acclaimed Kentucky Education   Reform Act requires schools to set high   expectation for students, but the same   can’t be said for teachers.
In 1989, the state Supreme Court   ruled the state’s public school funding   mechanism unconstitutional. It ordered   the General Assembly to start   from scratch in redesigning more equitable   funding. The result was the passage   of the 1990 reform law, which has   as its premise the belief that all students   can learn at high levels. While the   law was far from perfect, it did succeed   in focusing Kentucky’s attention on   school equity and student achievement.
But the law has failed to address   teacher quality. There are two major   reasons:
- Kentucky’s teachers union has used its political clout to successfully hold off attempts to toughen teaching standards.
- Lawmakers and bureaucrats have not had access to reliable data about the teacher workforce from which to make sound policy decisions.
These circumstances   prompted us to want to   look more closely at the   issues involved with   teacher competency and   quality in our state. This   meant also looking at the   ways in which teachers   are trained and whether   their roles as teachers   are adequately matched   with their preparation.
In reporting this series,   “The Learning Gap:   High Expectations for   Students, Low Standards   for Teachers,” education   reporters Holly E. Stepp,   Linda B. Blackford and I   documented that   Kentucky’s teachers are   able to receive a passing   grade on their certification   exams despite producing   some of the lowest   scores among similar   test-takers throughout   the South.
One invaluable resource   we used for developing   our reporting on this series   was an electronic database of information   about teachers that helped us to   gauge their quality. We supplemented   this analysis of data with interviews   with policymakers, school administrators   and teachers.
Arranging the interviews was easy;   getting access to the necessary data   was not. State education officials balked   at turning over what they considered   confidential information, including   details about teachers’ academic   records and their scores on state-mandated   certification exams. And some   data that we would like to have had,   such as tests taken by teachers before   the state began using PRAXIS and solid   demographic data such as race and   gender, simply didn’t exist in an electronic   form that could be trusted.
The newspaper argued in a series of   letters and telephone calls with the   attorney for the Kentucky Department   of Education’s Professional Standards   Board that the information, as a whole, should not be regarded as confidential.   Because Kentucky’s school reform   law evaluates schools largely based on   students’ test scores, why shouldn’t   teachers be judged by their scores on   teacher certification tests? Kentucky   children are tested annually on what   the state has decided they should know.   Each year, as the reform law spells out,   the bar for schools is raised slightly   higher. If a school declines beyond a   certain point, it faces sanctions.
But for Kentucky’s teachers, the standards   haven’t changed.
Getting the information we thought   we needed to do our own analysis of   teacher preparedness and performance   required two months of negotiation   with the Standards Board’s attorney.   In the end, state officials released the   scores on the three core PRAXIS certification   tests required in Kentucky and   more than two dozen specialty tests   used for individual teachers. They also   released to us data about the college   each teacher attended, deleting names   or demographic information that might   have allowed us to identify any individual.   In a second database, the state   released the PRAXIS test scores and   told us the school district—but not the   specific school—where each teacher   went to work after becoming certified.   The state’s data expert also changed   the identifier for each person in the   separate tables to make it impossible   for us to match the tables and identify   individuals.
For the most part, this arrangement   worked. Only in some extreme examples—such as one teacher who took   the three required certification tests a   total of 22 times—was it easy to learn   where that person ended up working.   (That person never was able to pass all   three tests and is a teacher working   with “emergency certification” in a   county in the south central part of the   state, not far from where he or she   graduated from college.)
We asked for data that went back 10   years. But we soon discovered that the   state has very little electronic data about   teachers prior to 1994. For us to have   received hard copies of these records,   after the state blacked out the names   and other identifiable information,   would not have been worth the cost or   time it would have taken us to put it   into the electronic form we needed to   do our analysis. But we never faced   that hurdle since the state couldn’t   guarantee that the older data we wanted   existed in any form.
Because there is so little electronic   data about teachers hired before 1994,   it became impossible for us to draw   any solid conclusions about their qualifications   as teachers. And this was illuminating   because it meant state officials   couldn’t do that kind of analysis,   either. This is a significant problem   since more than half of our state’s   teachers have been in the classroom   for at least 15 years. This means they   never took the PRAXIS tests, which the   state began using during the 1990’s,   yet they still hold their lifetime certification.   These teachers will remain certified   despite never having taken any   kind of test or doing anything more   than attend to the annual required   professional development, which we   found to be mediocre in many places.
Now, having said that, all Kentucky   teachers must get master’s degrees,   but they have 10 years in which to do   that. Many experts with whom we spoke   for this story questioned the value of   earning a degree that can be stretched   over that length of time. Essentially,   what our search for information let us   know is that the older age of many of   Kentucky’s teachers is a potential barrier   if one is trying to find ways to gauge   their qualifications.
Do these circumstances mean that   many of Kentucky’s teachers aren’t   qualified for their jobs?
The answer is that no one really   knows, because nobody had ever tried   before to compile the data and analyze   it. Looking at the individual test scores   by using Visual FoxPro, a computer   program that lets us query information   from a database, we were able to provide   a glimpse at some of the information   that is useful when measuring   teachers’ competency:
- We could let readers know how many teachers passed certification tests the first time by combining the scores we had with other published data on passing scores.
- We could let readers know how many teachers failed the first time and in which colleges they were trained. (Such comparisons had not been made public before our series appeared.)
We found that in Kentucky, to no   one’s surprise, people tend to go to   college near where they were raised,   and then they often return to their   hometowns to teach. This means that   students who came out of weak school   systems tend to return to them. And   this sets into motion a destructive cycle   that makes academic excellence harder   to achieve in these areas of the state.
By also analyzing the test scores,   using a technique developed by an   economics professor at Carnegie   Mellon University, we learned how   little is actually necessary to pass   Kentucky’s teacher certification test.   In fact, teachers were being allowed to   pass the three basic tests and 18 specialty   exams we reviewed by answering   anywhere from 35 percent to half   of the questions correctly.
The state agency that oversees   teacher training told our reporters that   it intends to raise the bar for passing   teacher certification tests and expects   to release specifics of its plan early this   year. Officials say they will require new   Kentucky teachers taking the PRAXIS   tests to score at least as high as those   elsewhere in the South, a region of the   country in which these test scores, in   general, are lower than in other areas.
So the state is faced with several   dilemmas. Is it enough to want its teachers   simply to score as well as they do   elsewhere in the South? Or should the   bar be set higher for them, as it is for its   students? This remains a delicate balancing   act for state officials, as our   reporters discovered: raise the bar too   high, too quickly, and not enough teachers   will pass. That leaves students without   teachers.
What we hope our series helped   readers and policymakers understand   is why quality matters in terms of how   teachers do their jobs and how being   lax in terms of letting teachers pass   tests at lower thresholds might jeopardize   children’s learning. Certainly our   series did succeed in getting the state’s   bureaucrats and policymakers to talk   seriously about the need for better   teacher training and higher qualifications.   The day after our series began   appearing in the newspaper,   Kentucky’s Education Commissioner   released the outline of a plan to improve   education. Teacher training and   quality were focal points of many of his   initiatives. Since these had not been   top agenda items before, the education   establishment was caught by surprise.
And then in the middle of January,   the newspaper learned the Governor   was planning to create an 18-member   task force to examine teacher training   issues. One state lawmaker said teacher   training has been a “hot topic” for   some time, but that the paper’s series   heightened awareness of the low standards   set for teachers even more.
That, to some degree, was our mission   in setting out to do this series. We   wanted to take an aspect of education   to which enough attention had not   been paid, shine a light on what could   be learned by carefully analyzing data,   and provide some vivid examples—through our reporting—that would   exemplify the harm that can occur if   the problem is left untreated. It took a   lot for us to get the numbers we needed   out of the state bureaucracy, but our   efforts to do so paid off handsomely   when the articles ran and, for the first   time, people were able to make connections   where none had been visible   before. 
Photo by Michelle Patterson, The Lexington Herald-Leader.
Photo by Michelle Patterson, The Lexington Herald-Leader.
   Linda J. Johnson is education writer   for The Lexington Herald-Leader.   Earlier this year she became the   paper’s Computer-Assisted Reporting   Coordinator. She began her involvement   with computer-assisted reporting   when she was health/environment   reporter for The Vindicator, a   paper in Youngstown, Ohio. 

 
                     
                    