For years, Maryland has been known for its excellent public schools. That doesn’t mean they are all of equal quality, but the state has been ranked No. 1 for five straight years by Education Week, which, if you put any stock in rankings, is the best on this subject because of the multiple indicators it considers.

And for years the Montgomery County public school system has implemented a nationally renowned teacher-evaluation system that is led by teachers and that works with teachers who need help to improve but also terminates teachers who are consistently ineffective in the classroom.

Now, according to this article by my colleague Lynh Bui, Maryland’s State Department of Education has rejected new teacher-evaluation proposals from Montgomery and Frederick counties and seven of the state’s 24 other county school systems. The reason: The proposals, officials said, don’t comply with state law, federal education reforms or both.

Detailed explanations for the rejections will come this week, but the big issue is likely to be the use of student standardized test scores as part of a teacher’s evaluation. The Obama administration has pushed states to rewrite their evaluation systems to include test scores (among other things), despite a consensus among assessment experts that these tests are not designed for this purpose and the results are unreliable.

In fact, some of the new evaluation systems linked to test scores are so wacky that teachers wind up being evaluated by scores of students they don’t even teach. Scores are plugged into formulas that purport to measure the value a teacher adds to a student’s learning, and that are supposed to be able to factor in all outside elements that could affect student performance, such as being hungry, tired or suffering from post-traumatic stress disorder. (“Supposed to be” is the key phrase in that sentence.) This high-stakes use of test scores is leading to “teaching to the test” as well as a growing number of cheating scandals.

It is true that most teacher evaluation systems in the country needed to be revised, but there are effective ways to do it, and non-effective and even destructive ways. “Value-added systems” too often wind up giving good grades to bad teachers and bad grades to good teachers.

In recent years, the Obama administration put pressure on states to include test scores in teacher evaluation through its Race to the Top competition — in which states vied for federal funding with promises to implement policies preferred by Education Secretary Arne Duncan — and through waivers the Education Department granted from restrictive mandates of No Child Left Behind, granted to those states that promised to go ahead with Duncan-supported policies.

Maryland officials won Race to the Top money in two separate rounds of the sweepstakes. In 2010 it won $250 million for promising to reform elementary and secondary schools, and in 2011 it won up to $50 million in the early education round of Race to the Top.

But in Maryland, two public school systems — Montgomery and Frederick — said they didn’t want Race to the Top money because their educational leaders did not believe that the promised reforms were smart. School systems that accepted the money had to link 50 percent of teacher-evaluation systems to student achievement, which basically means standardized test scores. However, Montgomery and Frederick were not bound by that but rather to a state law — passed by legislators who don’t understand assessment — that requires that student achievement be a “significant” factor in the evaluation.

What does “significant” mean? The state will soon explain its definition, but the bigger question is why Maryland officials think it makes any sense to require Montgomery County to change a system that is known to work.

The system used for years to evaluate teachers in Montgomery County did not put any absolute weight on test scores, although they could be used to look at trends about a teacher’s effectiveness.

In recent years as school reform leaders pushed to force states to link test scores to evaluations, Montgomery County officials did not broadcast the fact that their system didn’t require that tests be used. In fact, last year, Duncan said in a meeting with Washington Post reporters and editors that he thought Montgomery County did use test scores. When I asked him who told him that, he said Jerry Weast, the former superintendent of schools. When I asked Weast if Duncan was right in saying that test scores are used, or I was right in saying they weren’t, he said, “You are both right.”

Actually, the old system did not assign a test score any definitive weight, and the new system didn’t either; it gives principals the option to use scores from the Maryland School Assessment as one of many pieces of data that can be used to evaluate a teacher. Neither did the new system that Montgomery County submitted that was just rejected by state officials.

Bui quoted Superintendent Joshua P. Starr as saying: “I’m disappointed that the strengths of the Montgomery County Public Schools system are being compromised as a result of this decision. We have clearly shown over the years that a collaborative approach to teacher evaluations and support that also uses student achievement data sets the stage for improvement of student achievement.”

Starr has a right to be disappointed. All of the superintendents in Maryland should be. They surely know that linking test scores to teacher and principal evaluations is bad practice. So why don’t  the superintendents in the top-ranked state say so publicly? Where are Starr’s colleagues on such an important issue?

Something is wrong in the state of Maryland.