How to Make Short, Frequent, Unannounced Classroom Visits Work 

In this Kappan article, Kim Marshall addresses what he calls a “major blind spot” among educational researchers and reformers: the fact that most evaluative visits to teachers’ classrooms are announced in advance. Teachers, quite understandably, take their performance up a notch for these infrequent, high-stakes inspections, and students tend to behave better too. Could this be the real explanation for rampant grade inflation in teacher evaluations across the nation, as documented by the New Teacher Project’s 2009 Widget Effect study (23,332 Chicago public-school teachers rated Superior, 9,176 Excellent, 2,232 Satisfactory, and only 149 Unsatisfactory over a recent five-year period, with a similar skew in several other districts)?

Some educators defend announced observations. “I want to see what the teacher is capable of,” said one former superintendent. But is the teacher’s glamorized lesson representative of what students are getting day to day? asks Marshall. “I can see right through the dog-and-pony show,” said a seasoned principal. But can the principal document his hunch? “I need that pre-observation conference for feedback on my lesson planning,” said a teacher. But how helpful is discussing a lesson plan once a year, especially if it’s not representative of usual preparation?

Why do so many school administrators give credence to lessons that are clearly atypical? Marshall lists some possible reasons:

• Avoidance – A plausible lesson allows the administrator to skirt difficult conversations and the hard work of documenting and supporting a mediocre or ineffective teacher.
• The fundamental attribution error – School leaders tend to assume that seeing what looks like a good teacher means that good teaching is going on day to day.
• Infrequency – The conventional process (pre-observation conference, full-lesson observation, detailed write-up, and post-observation conference) is so time-consuming for administrators that teachers are evaluated only once or twice a year, and it would clearly be unfair for such infrequent visits to be a surprise. 

“That’s why districts, even without union insistence, have administrators schedule their formal observations in advance,” says Marshall.

This time-honored dynamic might seem benign and unavoidable, but it has serious consequences. Effective teachers don’t get authentic praise. Mediocre teachers don’t get targeted coaching and support. And all too many ineffective teachers are not held accountable. “To put it bluntly,” Marshall says, “an evaluation process that relies on announced visits is inaccurate, dishonest, and ineffective… This contributes directly to America’s widening achievement gaps, since students with any kind of disadvantage desperately need effective teaching.”

But what’s the alternative? Marshall argues that a number of principals are already implementing a better approach. Here are three layers of change:

• Changing the structure – Classroom observations shift from being announced, infrequent, and full-lesson to unannounced, frequent, and short:

• Unannounced – “There’s only one way a principal can look parents and other stakeholders in the eye and assure them of the quality of day-to-day teaching,” says Marshall: “regularly dropping into all classrooms without advance notice.” 
• Frequent – He believes ten visits to each teacher each year is enough to get a representative sampling of the teacher’s work in different subjects or with different classes and at different points in lessons and the school day, week, and year. In most schools, this means making about two or three classroom visits a day (multiply the number of teachers supervised by 10 and divide by the days in the school year). 
• Short – Watching a 5-15-minute video of a teacher in action is enough to satisfy virtually all educators that a lot happens in a classroom in a short period of time and there are always several “teaching points” to take up afterward. Of course, pre-observation conferences are not possible with these visits, cutting down on the time required for each.

“When observations are unannounced, frequent, and short, the supervision dynamic changes dramatically,” says Marshall. “School leaders have a much better sense of what’s going on in classrooms, and teachers find the process less stressful and believe their bosses get what they’re doing with students. In addition, administrators’ increased presence in classrooms, corridors, and stairways prevents many problems.”

• The human element – To have an impact on teaching and learning, Marshall argues that follow-up needs to be:

• Face to face – After each classroom observation, administrators should make a point of having a brief, informal conversation with the teacher, ideally within 24 hours. “Every time administrators talk with a teacher after a short visit, they learn something new,” says Marshall, “widening their observational window and improving trust.” 
• Humble – A winning strategy is for administrators to have these post-observation chats in teachers’ classrooms. “Being on the teacher’s turf changes the power dynamic,” says Marshall, “and there’s the additional advantage of seeing student work, curriculum artifacts, and other reminders of what was happening during the observation.” It’s also important for administrators to slow down and be good listeners; teachers have a lot on their minds.
• Honest – Making lots of short, unannounced visits, administrators will see good teaching, but they will also stumble across some mediocre and ineffective practices. If they don’t step up to the plate and address them, the whole process is a waste of time. “Difficult conversations” are difficult. Superintendents can help by having principals role-play with each other to improve their skills.
• Linked to teacher teamwork – One-on-one conversations with teachers after short observations should be linked to teacher teams’ work on curriculum planning and analysis of student assessments. “This sharpens administrators’ ‘eye’ and boosts the power of teacher teamwork,” he says.

Marshall disagrees with giving teachers initial feedback via e-mails, checklists, and electronic programs. “This kind of one-way feedback is superficial, bureaucratic, annoying, and highly unlikely to make a difference,” he says. The same goes for rating each drop-in on a 4-3-2-1 or Satisfactory/Unsatisfactory scale. “This increases the teacher’s anxiety level and is the opposite of good coaching.” Of course if a teacher is ineffective and not responding to support, the process needs to become more formal: longer classroom visits (unannounced, of course), an improvement plan, a timeline for improvement, and possible dismissal.

• Management details – Short, unannounced classroom visits are not the same as “managing by wandering around,” says Marshall. These observations need to be:

• Systematic – If administrators don’t keep a paper or electronic checklist of all staff being visited, there’s the tendency to lose track, not be equitable about visits and feedback conversations, and avoid certain classrooms. It’s very helpful to have a goal for visits (perhaps two or three a day) to keep up the pace. And if there is more than one administrator in the building, it’s important to divide up the workload.
• Documented – Marshall believes that each face-to-face feedback conversation should be followed up with a very brief write-up for the record (shared, of course, with the teacher). The sequence is important to building trust and dialogue: visit, conversation, then write-up. 
• Linked to end-of-year evaluations – “Yes, these short observations count,” says Marshall. “They are an artful blend of supervision, coaching, and evaluation, supporting teaching and learning and, each time, giving the administrator a few more pieces of the puzzle for the teacher’s summative evaluation.” The best way to capture end-of-the-year performance is by using a rubric, with each teacher sharing his or her self-assessment and comparing, page by page, with the administrator’s tentative evaluation and debating any differences in light of the evidence. 

Virtually all educators agree that ten short, unannounced classroom visits followed by feedback conversations give a much more accurate picture of a teacher’s performance than one or two dog-and-pony shows. If administrators handle them well, the effect can be dramatic.

The logic of this approach is compelling, says Marshall, but some districts are implementing a hybrid model, with announced and unannounced visits. “This sounds like a sensible, middle-of-the-road compromise,” he says, “but it has a fatal flaw: If principals continue to spend four hours or more on each traditional observation cycle and don’t get relief from other responsibilities, they simply won’t have time for more than one or two short observations – and that isn’t nearly enough for teachers to trust the process and for administrators to get a true sense of what’s going on in classrooms. The result will be exhausted and cynical school leaders and no improvements in teaching and learning.

“Let’s face it,” Marshall concludes: “Announced, infrequent, full-lesson classroom visits are bogus. Half-measures won’t work. We must make a clean break with the past and use an approach that will win teachers’ trust, provide continuous feedback on their work, fuel teacher teamwork, and culminate in accurate end-of-year evaluations.”

“Let’s Cancel the Dog-and-Pony Show” by Kim Marshall in Phi Delta Kappan, November 2012 (Vol. 94, #3, p. 19-23), http://www.kappanmagazine.org; Marshall can be reached at kim.marshall48@gmail.com.