It's Time to Rethink Teacher Supervision and Evaluation

It's Time to Rethink TeacherSupervision and EvaluationThe process by which most teachers are supervised and evaluatedis inefficient, ineffective, and a poor use of principals' time, arguesMr. Marshall. It needs to be drastically streamlined and linked to abroader strategy for improving teaching and learning.

BY KIM MARSHALL

PRINCIPAL boasts that he spends two hours a day in classrooms. And it's true - he really does visithis school's 1 7 teachers daily, chatting with students and occasionally chiming in on a lesson. Butwhen teachers areasked what kind offeedback they get,they say the principal A -.

rarely talks to themabout what he sees . -

when he strolls through their classes.

A principal gets complaints from L Iseveral parents about a history teach-er's problems with discipline but is sooverwhelmed that she rarely visits hisclassroom. When she does her requiredobservation of his class, she sees a care-fully planned lesson featuring an elab-orate PowerPoint presentation and well-

KIM MARSHALL was a teacher, a centraloffice administrator, and a principal in theBoston Public Schools for 32 years. He nowworks as a leadership coach for new prin-cipals in New York City and Newark, NJ.;teaches courses on instructional leadership;and publishes The MarshallI Memo, a weeklydigest of ideas and research on K-12 educa-tion (www.marshallmemo.com). He wish-es to thank Roland Barth, John King, SandyKleinman, Jay McT7ghe, Douglas Reeves, JonSaphier, Mike Schmoker, Rhoda Schneider,and Athie Tschibelu for their important con-tributions to this article.

Illustration by Brenda Grannan JUNE 2005 727

behaved students. The principal feels she has no choicebut to do a positive write-up of this lesson and give theteacher a satisfactory rating.

* * *

A principal spends four entire weekends in April andMay completing teacher evaluations just before the dead-line. He puts the evaluations into teachers' mailboxes witha cover note attached that reads, "Please let me know if youhave any concerns and would like to talk. Otherwise, signand return by tomorrow." All the teachers sign, nobody re-quests a meeting, and there is no further discussion.

A well-regarded veteran teacher hasn't been evaluated infive years and rarely sees the principal in her classroom. Shetakes this as a compliment - her teaching must be "okay."And yet she feels lonely and isolated with her students andwishes the principal would pay an occasional visit and tellher what he thinks.

* * *

A sixth-grade teacher has good classroom managementand is well liked by students and parents, but his studentsdo poorly on standardized tests. A new principal mentionsthe disappointing scores, and the teacher launches into alitany of complaints: he always gets the "bad class," mostof his students come from dysfunctional families, and he'stired of being asked to "teach to the test." Later that day, theunion representative officiously tells the principal that shecan't mention test results in a teacher's evaluation.

* * *

A principal observes an elaborate hands-on math lessonin a veteran teacher's classroom and notices that the teach-er is confusing the terms mean, median, and mode. The prin-cipal notes this error in his mostly positive evaluation, and,in the post-observation conference, the teacher suddenly be-gins to cry. Ten years later, at his retirement party, the prin-cipal asks the teacher what lesson she took away from thisincident. "Never to take a risk," she replies.

* * *

The theory of action behind supervision and evaluationis that they will improve teachers' effectiveness and there-fore boost student achievement.' This assumption seems log-ical. But the vignettes above raise a troubling question: whatif the theory is wrong?This article takes a close look at thispossibility and explores an alternative theory of action.

WHY DO SUPERVISION AND EVALUATIONOFTEN MISS THE MARK?

I believe there are 10 reasons why the conventional su-pervision and evaluation process is not an effective strate-

gy for improving teaching and learning.1. Principals evaluate only a tiny amount of teaching.

If a teacher has five classes a day, that's 900 periods eachschool year. A principal who formally evaluates a teacherfor one full class period a year (a fairly typical scenario)sees this proportion of the teacher's lessons:

In this case, the principal evaluated 0.1% of the teacher'sinstruction. The other 99.9% of the time, the teacher wasworking with students unobserved. Even if the principalmade three full-class evaluation visits a year, as requiredby some districts, that would still leave the teacher alonewith students 99.7% of the time. No matter how observantand well trained the principal is, no matter how comprehen-sive the evaluation criteria are, and no matter how detailedthe feedback is afterwards, this is ridiculously thin super-vision of the school's most important employees. Principalswho spend this little evaluative time in classrooms are ba-sically bluffing, hoping that teachers will think they knowmore than they really do. Without expensive increases inadministrative staffing -politically impossible in most dis-tricts - the amount of time principals spend formally ob-serving each teacher is not going to change. Let's face it:teachers are on their own most of the time, and our schoolsdepend heavily on their competence and professionalism.

2. Microevaluations of individual lessons don't carry muchweight. Many school districts try to compensate for how littletime principals spend in individual classrooms by requiringextremely thorough evaluations of lessons that are formallyobserved. Administrators are asked to script everything theteacher says and write a detailed account of exactly what hap-pened in the class. A perceptive and well-trained principalcan see a lot in a single lesson and give the teacher copiousfeedback on classroom management, student engagement,"accountable talk," clarity, momentum, wait time, bulletinboards, and so forth. But these elaborate write-ups don't meana lot to most teachers; they know how little the principal sees

728 PHI DELTA KAPPAN

of their day-to-day struggles, curriculumplanning, grading, work with colleagues,parent outreach, professional growth,and routine duties. Even if the evaluationis complimentary, it usually gets filed in - Xa nanosecond. Except in extreme (andquite rare) cases when a principal givesan unsatisfactory rating, evaluation is a Skillpro forma process that has very little in-fluence on what teachers do on a daily X X Tbasis.

3. The lessons thatprincipals evalu- . Late are often atypical.The only way that -microevaluating lessons can give an ac- fcurate picture of a teacher's overall class-room performance is if the observed les- :sons are truly representative. But this isoften not the case. When teachers haveadvance notice of an evaluation, they The Skilful L

can present a glamorized lesson for the the fied of e,principal's benefit. Even if they don't,the presence of a top-level authority fig-ure in the classroom usually reduces dis-cipline problems and results in a moreorderly lesson than students generallyexperience. These two factors can workin teachers' favor, giving the principal an unrealistically pos-itive view of their teaching. You'd think that principals wouldbe wise to these dynamics, but they are often so stressedand overwhelmed that they play along, treating clearlyatypical teaching as typical. When this happens, teachersget an unfortunate message: it's okay to do "special" teach-ing when the principal visits and "ordinary" teaching forstudents the rest of the time.

Evaluation visits can also distort reality in.a negative way:some teachers get so nervous when the principal arrives thatthey go to pieces. This is every teacher's nightmare -onescrewed-up lesson and the other 99.9% of the year will bepainted with the same evaluative brush.

Surely the principal has other sources of information tocorrect egregiously off-target observations, including informalvisits, quick impressions of teachers interacting with students,parent comments, colleagues' impressions, and gossip. Butthese time-honored sources of information, even when ac-curate, aren't "admissible" in official evaluations. Principalshave little choice but to go by the book and use the infor-mation from formal evaluation visits, even when it's bogus.

4. Isolated lessons give an incomplete picture of instruc-tion. Although the lesson is the fundamental building blockof teaching, it's only a small part of a teacher's effort to in-

j b 2.J .UJJ

Eel Leader

.earn field-tested tools and strategiesto help education leaders move teachersFrom mediocrity toward excellence.

Jse these procedures and models forwriting observation reports, summaryevaluatioris, and improvement plans.

Leader without question is a seminal contribution toduceilonal leadership and supervision."- Je4y Patterson, Professor of Ed. Leadership, Univ._of Alabama

Research for Better Teaching, Inc.One Acton Place. Acton. MA 01720Phone (978)263-9449 * Fax (978)263-9959Web wwwrbteach.com * Email iinfofrbleach.com

spire students and convey knowledge and skills. To grasp thebigger picture, a principal needs to know more: What cur-riculum unit is this lesson part of? What are the unit's "bigideas" and "essential questions"? How does this unit alignwith state standards? How will students be assessed? Prin-cipals may try to ferret out these missing pieces by askingfor lesson plans and conducting pre- and post-evaluationconferences with the teacher, but evaluations are still tiedto the lesson that was observed.

This is a shame, because it's.impossible to teach moststate standards in a single lesson; it's a huge leap from big-picture goals like "understanding number sense" to plan-ning a single lesson. Unit plans, which describe a teacher'sgame plan for teaching skills and concepts over a three- tofive-week period, tell far more about whether instruction iscoherent and aligned. But principals rarely see unit plans orthe assessments that teachers give at the end of their units.

5. Evaluation almost never focuses on student learning.In virtually all school districts, teacher unions have been suc-cessful in preventing their members from being evaluatedon whether students actually learn what's being taught. Unionsare right to object to accountability on norm-referenced tests,since these assessments are not designed to be "instruction-ally sensitive."2 Before-and-after, "value-added" assessments

JUNE 2005 729

are better, but even their most fervent advocates don't thinkit's fair to use them to evaluate a teacher after only a yearof instruction.

Does this mean that principals have no way of evaluat-ing teachers on whether students are learning? Surely a prin-cipal can get a sense of how much students are picking upby walking around classrooms, looking over their shoulders,and asking them probing questions. But this approach hasthree problems. First, many principals are required to pro-duce detailed narratives after each evaluation visit and can'twalk around and write furiously at the same time. Second,even if principals manage to check in with a few students dur-ing classroom visits, it's hard to tell whether the whole classunderstands the lesson that day - let alone a few weekslater. To really know if teachers have been successful, prin-cipals need to see students' scores on good unit assessments- which they almost never do. Third, even if principals canget their hands on interim assessment results, such evidenceis not admissible in evaluations.

So principals have little choice but to focus on teachingperformances versus learning results, on chalkboard razzle-dazzle versus deep understanding, on beautiful bulletin boardsversus demonstrated proficiency. Constrained by the super-vision/evaluation process, principals overmanage the occa-sional lesson and undermanage the bigger picture of wheth-er teachers are truly making a difference in student learn-ing.

6. High-stakes evaluation tends to shut down adult leam-ing. Even though many teachers don't respect the evaluationprocess, it still makes them nervous. Their collective bargain-ing agreements may provide good protection, but teachersharbor irrational fears that every time the principal walksinto their classrooms, clipboard in hand, their jobs are onthe line. Formal evaluations raise the level of tension andanxiety and make it more difficult to admit errors, listen, andtalk openly about areas that need improvement. Any timeevaluative comments are put in writing, the parties involvedtighten up: the principal is less likely to tell the whole storyfor fear of facing a grievance, and the teacher is less likelyto talk about how things are really going. In all too manyevaluative interactions, teachers put on their game face andget through the process with as little authentic interactionas possible. The principal owns the feedback, not them.

This kind of process destroys a golden opportunity forprofessional growth. The real challenge of supervision andevaluation is to activate (or amplify) a supervisory voice in-side teachers' heads that will guide them in their work withstudents. Conventional supervision and evaluation seldomaccomplish this goal. In fact, the exact opposite may occur,with teachers waiting nervously for their principal to judge

them and putting up a wall of resistance to any criticism.Where do teachers go for helpful feedback on their teach-ing? Usually they turn to a colleague, a spouse, a familymember, students, parents - or nobody.

An unintended consequence of this whole dynamic isthe growth of a certain emptiness in the professional rela-tionship between teachers and school leaders. If principalsare rarely in classrooms, it's hard to have meaningful profes-sional conversations with teachers. And if principals aren'tsetting the tone, it's less likely that assistant principals, teamleaders, department heads, and colleagues will have seri-ous conversations about teaching and learning. This kindof instructional vacuum can result in faculty lounge con-versation dominated by topics outside of the school, gos-sip, and funny - and not-so-funny - stories about kids.,

7. Supervision and evaluation reinforce teacher isolation.One of the American principal's toughest challenges is coun-teracting two tendencies prevalent in our schools: teachersnot working with their colleagues and the 'educator's ego-centric fallacy" - I taught it, therefore they learned it. 4 Infar too many schools, teachers who teach the same subjectsat the same grade level don't work together, missing outon the synergy of collaboration and wasting precious timereinventing the wheel. Because principals evaluate teach-ers in private meetings and confidential documents, evalua-tion reinforces this isolation and is rarely a vehicle for get-ting teachers to talk to one another, which detracts fromteachers' sense of responsibility to their grade-level or de-partment team.

Evaluation is also an ineffective tool for countering ournatural tendency to assume that if something is taught (i.e.,explained or demonstrated), it is automatically learned. 5 Be-cause the supervision and evaluation process doesn't fo-cus on team curriculum planning, assessment, and studentlearning, it doesn't prod teachers to emerge from their iso-lation and reflect with their colleagues on what they needto change in order for more students to succeed. Withoutthis impetus, teachers gravitate toward the default setting:self-contained, activity-centered lessons or marching throughthe textbook.

8. Evaluation instruments often get in the way. Good teach-ing is extremely complex and challenging, and research tellsus there is more than one way to get students to learn. It takesexperience and savvy for a principal to grasp the subtletiesof a classroom; it's even more demanding for a principal tocapture them in writing; and it's really challenging to criti-cize a teacher's performance in a way that is heard. Someprincipals are good at all three - observation, write-ups,and 'difficult conversations." Unfortunately, many princi-pals are not, and the training needed to bring them up to

730 PHI DELTA KAPPAN

III

II

III

II

i

iI

II

i

speed is woefully lacking. The legendary klutziness of schooladministrators has motivated unions to work overtime tonegotiate "principal-proof" evaluation formats and proce-dures to protect their members from unfair evaluations. Dis-tricts, on the other hand, push for evaluation tools that makeit possible to build a case to dismiss incompetent teachers.The resultant evaluation tools are rarely conducive to fos-tering an honest, open, and pedagogically sophisticated dia-logue between principals and teachers.

9. Evaluations often fail to give teachers 'judgmental"feedback.This seems like an odd statement, since all eval-uations judge teachers. But many evaluation instrumentsallow principals to fudge teachers' general status with anoverall "satisfactory" rating and a lot of verbiage. Theseevaluations don't tell teachers where they stand on clear-ly articulated performance standards, don't give clear di-rection on the ways in which teachers can improve theirperformance, and don't answer the question teachers real-ly care about (and often dread): How am I doing?This kindof evaluation is unlikely to motivate a mediocre teacher toimprove - or spur a good teacher on to excellence.

10. Mostprincipals are too busy to do a goodjob on su-pervision and evaluation. Discipline and operational dutiesare so insistently demanding that teacher evaluation oftendisappears from principals' calendars until contractual dead-lines force them to get serious.6 When evaluation crunch timearrives, principals fall into three types - saints, cynics, andsinners. The saints go by the book, and evaluation consumestheir lives for weeks at a time. I know a principal who rou-tinely spends eight to 10 hours on each teacher evaluation:pre-observation conference, lesson observation, write-up(like a little term paper every Saturday, she says), and post-observation conference. Principals who choose to com-mit this amount of time (or are required to do so by theirsuperiors) have no alternative but to shut themselves in theiroffices for days at a time - or spend evenings, weekends,and vacations at their desks at home. Ironically, this reducesthe amount of time the saints spend in classrooms doinglow-key supervision -coaching, encouraging, and gentlecorrection.

The second type of principal heaves a sigh, sits down atthe computer, and bangs out the required evaluations asquickly as possible. Administrators in this category havegrown cynical about the evaluation process and don't be-lieve their write-ups will produce better teaching and learn-ing, but they feel they have no choice but to do them.

The third, more daring, group of principals simply don'tdo evaluations (or evaluate only the occasional egregious-ly ineffective teacher). These sinners ignore contractual re-quirements and dare the system to catch them. Since evalu-

ation is in the same category as a trip to the dentist for manyteachers, they tend not to complain if their principal "for-gets" year after year. And principals' superiors are oftennone the wiser - or choose to wink at these omissions.

So here's the question: are the saints, who spend hourson each evaluation, more effective at improving teachingand learning in their schools than the cynics and the sinners?Shocking as it may seem, the answer in many cases is no.This is because the conventional supervision and evalua-tion process is not the best way to truly change what hap-pens in classrooms. Principals need a better way to observe,support, and judge teachers - a way that is more accurateand time efficient and more closely linked to an effectivestrategy for improving teaching and learning.

LINKING SUPERVISION AND EVALUATIONTO HIGH STUDENT ACHIEVEMENT

I've argued that the theory of action behind supervisionand evaluation is flawed and that the conventional processrarely changes what teachers do in their classrooms. Here isan alternative theory: The engine that drives high studentachievement is teacher teams working collaboratively to-ward common curriculum expectations and using interimassessments to continuously improve teaching and attendto students who are notsuccessful. Richard DuFour, MikeSchmoker, Robert Marzano, Douglas Reeves, Jeffrey How-ard, Grant Wiggins, Jay McTighe, and others believe thatthis approach is a critical element in high achievement. Iagree, but with a proviso: if a school adopts this theory, itmust change the way teachers are supervised and evaluat-ed. If it doesn't, the principal won't have the time, energy,and insight to get the engine started and monitor it duringeach school year.

Why are the principal's time and focus so crucial? Be-cause teacher collaboration is countercultural in most Amer-ican schools and rarely happens without impetus and sup-port from outside the classroom. Principals are in the bestposition to provide the support, and rigorous state standardsand high-stakes tests can provide the impetus. Standards andtests present a common challenge (a common enemy, somewould say) that makes it easier for principals to get teacherteams to buy into working. toward ambitious, measurablelearning for students.

Of course principals still need to evaluate teachers everyyear or two, as required by moststates, and they also needto give honest and timely feedback to ineffective teachersand have the guts to fire them if they don't improve. Butthe essence of what I'm recommending is a shift away froma process owned by the principal, in which most of the en-

JUNE 2005 731

ergy goes into evaluating individual lessons, to a more dy-namic, informal process owned by teacher teams. To makethis happen, we need to shift:

* from periodically evaluating teaching to continuous-ly analyzing learning;

* from inspecting teachers one by one to energizing thework of teacher teams;

* from evaluating individual lessons to supervising cur-riculum units;

c from occasional announced classroom visits to frequentunannounced visits;

- from detailed scripting of single lessons to quick sam-pling of multiple lessons;

- from faking it with distorted data to conducting authen-tic conversations based on real data;

e from year-end judgments to continuous suggestionsand redirection;

* from comprehensive, written evaluations to focused,face-to-face feedback;

a from guarded, inauthentic conversations to candid give-and-take;

6 from teachers saying, "Let me do it my way" to every-one asking, "Is it working?";

* from employing rigid evaluation criteria to continu-ously looking at new ideas and practices;

* from focusing mainly on bad teachers to improvingteaching in every classroom;

* from cumbersome, time-consuming evaluations to stream-lined rubrics; and

- from being mired in paperwork to orchestrating school-wide improvement.

TWELVE STEPS TO LINKING SUPERVISION ANDEVALUATION TO HIGH SCHOOL ACHIEVEMENT

These shifts will not happen by themselves. To recoverfrom ineffective practices and to address widening achieve-ment gaps, principals might try the following 12-step pro-gram.

1. Make sure the basics are in place. These include timescheduled for teacher teams (grade-level teams in elemen-tary schools and subject-area teams in secondary schools)to meet on a weekly basis, preferably in uninterrupted 90-minute blocks; crystal-clear, end-of-year expectations forlearning that are aligned with state standards; common as-sessments, which can be written by teacher teams or pur-chased, to measure learning and diagnose needs at the endof each year and at intervals during the year; common ru-brics for consistently scoring student writing and open-endedresponses; and exemplars of student work at the advanced,

proficient, basic, and below-basic levels.2. Decide on the irreducible elements of good teaching.

For principals and teachers to communicate well aboutwhat's happening in classrooms, there must be a commonlanguage regarding the basics of effective teaching. Mostevaluation checklists are way too long to remember. A handyacronym for the five elements that every classroom shouldhave is SOTEL: safety- students feel physically and psy-chologically protected; objectives - the goals of the cur-riculum unit are evident; teaching- learning experiencesare skillfully orchestrated; engagement- students are lean-ing forward, involved in the learning process; and learning- there is evidence, either during the lesson or on follow-up assessments, that students have learned what was taught.

3. Systematically visit all classrooms on a regular basis.Principals need to be in classrooms frequently for a reali-ty.check on how things are going. But how frequent is "fre-quently," and how much time does a principal need to bein a classroom to see how things are going? The answersto these two questions are crucial because there's a directrelationship between the length of each visit, the numberof classrooms a principal can see each day, and the qual-ity of information that is gathered. Shorter visits mean theprincipal can cover more classrooms, but'visits that are tooshort yield superficial data.

Most principals make four types of classroom visits: 1)very brief, "showing the flag" appearances; 2) "walkthroughs"lasting a few minutes, with particular attention to studentwork on bulletin boards; 3) five- to 15-minute mini-obser-vations focused intently on teaching and learning; and 4)full-period, formal observations with detailed note taking.All four types of visits are useful, but as I have argued pre-viously,' the third type is optimal for teachers whose basiccompetence is not in question. Mini-observations allow theprincipal to fit as many as five substantive visits into a busyday, and, if the visits are unannounced and the principal isfocused and perceptive, they yield the most accurate dataon how well teachers are performing.

A principal who is self-disciplined about making threeto five mini-observations a day can get into all the class-rooms in a medium-sized school every two weeks, system-atically sampling the quality of teaching in chunks of timethat can be fitted into a busy day. Using this approach, aprincipal can take 12 to 15 "snapshots" of every teacher'sperformance in the course of the year and compile a "photoalbum" of each one's overall performance. The total timethe principal spends in each teacher's classroom is not muchlonger than that spent in the conventional evaluation modeldescribed earlier, but the accuracy of the information gainedis far superior.

732 PHI DELTA KAPPAN

There's an additional bonus for peripatetic principals: theyget to know students better and pick up information thatcan be useful in understanding learning problems, resolv-ing discipline situations, and talking with parents.8 Frequentclassroom visits also convey an unmistakable message toteachers: "You never know when I'll drop in, and I expectgood teaching to be going on whenever I do." If the prin-cipal sees something of concern (for example, a studentbeing publicly humiliated), it's time to shift gears to a for-mal reprimand or a traditional full-lesson evaluation.

4. Give teachers prompt, face-to-face feedback after everyclassroom visit. Teachers should not be left in the dark aboutwhat the principal thinks, and personal feedback is far pref-erable to sending e-mails or leaving notes in teachers' mail-boxes. In an informal, low-threat, private conversation, teach-ers are more likely to relax and engage in honest give-and-take about how things are going. These conversations gobest when the principal's feedback focuses on one or twospecific points - e.g., an appreciative comment about theway the teacher drew a shy student into the discussion, ora critical comment about the fact that the hands-on activ-ities weren't focused on the unit objectives. Follow-up talksare most effective when they happen within 24 hours: "Bet-ter 120 seconds of feedback the same day than a five-pageessay delivered a month later," says Douglas Reeves.9

In each of these follow-up conversations, principals shouldmake a point of asking about student learning: "How is theEgypt unit coming?" "What Fountas-Pinnell levels have yourlowest reading groups reached?" "How did the algebra testgo?" If a principal has established a trusting climate, a teach-er should be able to say, "My team just spent two weeksteaching the concept of borrowing, and the kids bombedon our quiz. Can you help us figure out what happened?"Teachers should know that their boss is keenly interestedin results and should be comfortable reaching out for sup-port.

5. Require teacher teams to develop common unit plansand assessments. The best way to ensure that teaching isdone right the first time (versus having to provide correc-tive instruction for substantial numbers of students after thefact) is to have teachers work in teams to plan each curric-ulum unit with the end in sight.)0 Before they dive into teach-ing, teacher teams should work backwards from the statestandards to identify clear learning objectives, decide on

,the big ideas and essential questions of the unit, draft as-sessments they will use to determine whether students havelearned what was taught, create a game plan and calendarfor instruction, and run the plan by the principal for feed-back.

The three- to six-week curriculum unit is an ideal chunk

2005LeadershipFor ClassroomAssessmentConferenceSeptember 28-30, 2005Hilton - Omaha, Neb.

Meet the educators who

b''4School-based

Teacher-led AssessmentReporting System

built Nebraska's School-based Teacher-led Assessmentand Reporting System (STARS), interact with nationalassessment experts, and learn about leadershipstrategies for classroom-based assessment andalternatives to a single state test.$250 registration fee includes meals and materialsfor all three days. For more information, visit www.Icaconference.com or write LCA Conference, 455South 11th St., Suite A, Lincoln, NE 68508.Ph. 402/476-8055.

of instruction for principals to supervise - far better thanan individual lesson. A principal who has reviewed a unitplan can check out alignment in classrooms, look at howkids are responding, suggest midcourse corrections, andask about student learning. Are examining unit plans andfollowing up with teachers time-consuming? Yes. Are theseactivities a better use of a principal's time than lesson write-ups that are ignored by teachers? Absolutely!

By far the hardest part of implementing this approachis getting teachers to plan together in the first place. Teach-ers in the U.S. are accustomed to autonomy, and it takes atenacious principal to foster this kind of collab6ration. It'sessential, though, because teams plan better than teachersworking solo, and teams generate stronger ideas, providebetter support, and increase the likelihood that the super-visory voice will be in each teacher's head as the unit un-folds.

6. Require teams to give common interim assessments.If formative assessments are of high quality - not just clonesof multiple-choice end-of-the-year tests - they can giveteachers-valuable insights into what students are learningand not learning.", It's vital for teams to meet after each unitor quarterly assessment to look at the results and collec-tively answer these three questions: What percentage ofstudents scored at the advanced, proficient, basic, or be-low-basic levels? In which areas did students do best, andwhere were they confused and unsuccessful? What is ourstrategy for addressing the weakest areas and helping stu-dents who are struggling?"2 A powerful enhancement to in-terim assessments is for teams to set SMART goals - Spe-cific, Measurable, Achievable, Results-oriented, and Time-

JUNE 2005 733

bound (for example, 85% of our first-graders will be read-ing at Level I on the Fountas-Pinnell scale by June) - atthe beginning of the year and to track prdgress each quar-ter.

Teacher ownership of this process is vital; it's better fora teacher to chair team meetings, ideally on a rotating basis,even if the principal has the time. Teachers need to have aclear mission for their meetings (experimentation, continu-ous improvement, and results), contractual time to score as-sessments, common planning time during or after the schoolday to analyze and discuss data, an outside facilitator (un-less there is unusually strong leadership within the team),and occasional drop-in visits by the principal to give sup-port and contribute ideas.

7. Have teams report on student learning after each unitor quarter. Lots of schools suffer from data overload andinsufficient analysis and follow-through. The principal canhelp teams crystallize their thinking by asking for a brief,informal report on the three questions above and on oneadditional question: How can I help? It's crucial that thesereports, which can be submitted either in person or in writ-ing, are low-stakes, nonthreatening, and nonbureaucratic.Teams shouldn't be bogged down in paperwork and mustfeel they can be creative, try new things, admit mistakes,and engage in an informal give-and-take about what's work-ing and what needs to be improved.

To summarize, let's contrast how a principal evaluatesa teacher using the conventional model with the processthat would be followed under the proposed model:

Conventional ModelPre-observation conferenceLesson planLesson observationEvaluation write-upPost-observation conference

with teacherOccasional walkthroughs

Proposed ModelTeacher team develops unit planTeam writes common assessmentsPrincipal gives feedback on these to

teamTeam meets during unit to share

ideasBrief principal visits to classroomsFace-to-face feedback to teachersTeam gives a common assessmentTeam analyzes unit learning resultsTeam reports results to principalPrincipal discusses results with team

8. Arrange for high-quality feedback on lessons for teach-ers. Once a principal has made the shift to short, frequentclassroom visits followed by face-to-face feedback and islooking at unit plans and successfully orchestrating teach-er teams to focus on student results, who will give teachersfeedback on full-period lessons? The principal won't havetime but might arrange for instructional coaches or otherteachers to do longer observations and follow-ups on les-

sons. Colleagues and coaches can give valuable feedbackto teachers, especially when their input is part of a "lessonstudy" process. But there's a potential problem with peersobserving one another- the culture of nice. It's hard togive critical feedback to people you eat lunch with everyday. Videotape is a better medium for taking an unsparinglook at a lesson. There's no better way to see the flaws inone's teaching (and appreciate the strengths) than to watcha videotape with a critical friend. Videotaping also requiresmuch less skill than writing up a lesson observation.

The goal of all supervision, whether it comes from theprincipal's short visits or from a more lengthy peer or videoobservation, is to foster a real openness to feedback, installthe supervisory voice in teachers' heads, and breed an acuteconsciousness of student learning results. We want individ-ual teachers and teacher teams to be thinking constantly aboutwhether students are learning and what can be done to getbetter results.

9. Create a professional learning culture in the school.Teachers and principals need preparation and support toimprove their skills at observing classrooms; giving frankand honest feedback; and assessing unit plans, tests, -anddata on student learning. The principal needs to be the"chief learner" in this regard, reaching out to the knowl-edge base and orchestrating study groups, article and bookgroups, peer observations, and lesson videotapes. The goalis to create a culture in which nondefensive analysis of stu-dent learning is "the way we do things around here."

The nine steps above could be carried out within mostcollective bargaining agreements. The last three would prob-ably require waivers or contract changes.

10. Use short observation visits to write teachers' finalevaluations. Dispensing with elaborate, announced eval-uations is a huge time-saver, and once a trusting climatehas been established, it's the ideal scenario. When I wasprincipal of the Mather School in Boston, teachers becameso comfortable with my short visits and personal feedbackthat virtually all of them agreed (via individual sign-offs withthe assent of the union representative) to allow me to skipformal observation visits entirely and use my 12 or so shortclassroom visits-with-feedback to write their final evalua-tions. (For teachers who were in danger of getting overallunsatisfactory ratings, I went by the book.) The LittletonPublic Schools in Massachusetts are in their second yearof a negotiated agreement that gives tenured teachers thechoice of being evaluated using the traditional approachor using evaluations based on at least 10 short visits."3

11. Include measures of student leaming gains in teach-ers' evaluations. Teachers could be asked to submit evi-dence of changes in student learning from the beginning

734 PHI DELTA KAPPAN

to the end of the year, using before-and-after assessmentresults or an analysis of portfolios and student work.

12. Use a rubric to evaluate teachers. Scoring guides arebeing used successfully to evaluate student writing and oth-er open-ended work, and a few school districts, includingAlexandria, Virginia, and the Aspire Charter Schools in Cal-ifornia, have begun to use them for teacher evaluation. Ru-brics have several advantages over conventional evalua-tion instruments: they are more clearly 'judgmental," forc-ing the principal to give the teacher clear feedback withrespect to a standard; they are more informative, tellingteachers where they stand on a 4-3-2-1 scale with a de-tailed description of what performance looks like at eachlevel of proficiency; they counteract "grade inflation," ifit's clear that very few teachers will, be at the advancedlevel; and they take much less time.

CONCLUSION

Let's return to the vignette of the teacher who wept afterbeing told that she had mistaught an important math con-cept. It's a true story; I was the principal. Looking back, I'vedone a lot of thinking about what went wrong in that sit-uation. The teacher was clearly putting on a special lessonfor my announced visit. Her nervousness about the bien-nial evaluation may have thrown her off her game, and thehigh-stakes nature of our conference undoubtedly contrib-uted to her feeling of devastation when, in her view, I played"gotcha." She had been working in isolation from otherteachers at her grade level and was probably more focusedon impressing me than on bringing her students to profi-ciency on a fair assessment. The lesson she drew from mycriticism - to "never take a risk" - seems like the wrongone, but given the supervision and evaluation process thatwe were using, it was understandable.

Had this teacher been working in the kind of profes-sional learning community I have advocated in this article,things might have gone differently. She and her teammateswould have planned the math unit together, caught the errorearly on, and figured out a classroom strategy for teachingthe concepts. The teachers would have been less concernedabout what I thought, if I happened to drop in on a lesson,than on whether the kids were getting it and how they woulddo on their interim assessments and on the rigorous Massa-chusetts math test. If I did catch a teaching error during aclassroom visit, I would have corrected it in an informalconversation. When their students did wel I on the end-of-unit assessment, the team teachers would have reported theresults to me and their colleagues with real pride - even,perhaps, with tears of a different kind.

If this scenario is to occur, some changes need to bemade. We need to streamline supervision and evaluationso that principals can spend their time doing what will makethe most difference: quickly and efficiently keeping tabs onwhat is really happening in classrooms, giving teachers con-stant feedback, making fair judgments about teacher per-formance, and getting teams invested in improving studentlearning and focused on results. Principals need to be ableto shape a creative, low-stakes, professional learning com-munity so that teacher teams can continuously improve theirstudents' chances of succeeding in a high-stakes world.

Principals are ideally situated to start this team-driven"engine of improvement" and keep it humming month af-ter month. A few maverick school leaders are already do-ing this kind of work on their own. Others need permis-sion from their superiors before they take the leap of faith,let go of the current model of supervision and evaluation,and launch a more powerful learning dynamic. I would ar-gue that liberating principals to do the right kind of workis one of the most important steps a school district can takeif it wants to close the achievement gap and get all studentsachieving at high levels.

1. The distinction between supervision and evaluation in this article isbetween formative and summative assessment of teachers' work, betweencoaching and judging.2. James Popham, "A Game WithoutWinners," Educational Leadership,November 2004, pp. 46-50.3. Personal communication with John King on 6 January 2005.4. Grant Wiggins and Jay McTighe, Understanding by Design, expand-ed 2nd ed. (Alexandria, Va.: Association for Supervision and CurriculumDevelopment, 2005).5. Graham Nuthall, "Relating Classroom Teaching to Student Learning:A Critical Analysis of Why Research Has Failed to Bridge theTheory-Prac-tice Gap," Harvard Educational Review, Fall 2004, pp. 273-306.6. To reduce principals' evaluation workload, Jon Saphier has suggestedadopting a four-year evaluation cycle in which each teacher gets an in-depth principal evaluation once every four years and rotates through otherkinds of assessment (e.g., peer evaluation, a study group, self-assessment)in the other three. Jon Saphier, How to Make Supervision and EvaluationReally Work (Acton, Mass.: Research for BetterTeaching, 1993).This ap-proach sounds promising, but with high rates of staff turnover, it is notwithout problems.7. Kim Marshall, "How I Confronted HSPS (Hyperactive Superficial Prin-cipal Syndrome) and Began to Deal with the Heart of the Matter," PhiDelta Kappan, January 1996, pp. 336-45; and idem, "Recovering fromHSPS (Hyperactive Superficial Principal Syndrome): A Progress Report,"Phi Delta Kappan, May 2003, pp. 701-9.8. Personal communication with John King on 6 January 2005.9. Personal communication with Douglas Reeves on 5 January 2005.1 0. Wiggins and McTighe, op. cit.11. Douglas Reeves, Accountability for Learning: How Teachers andSchool Leaders Can Take Charge (Alexandria, Va.: Association for Super-vision and Curriculum Development, 2004).12. Mike Schmoker, Results: The Key to Continuous School Improvement,2nd ed. (Alexandria, Va.: Association for Supervision and Curriculum De-velopment, 1999).13. For information on the Littleton, Mass., teacher evaluation contractlanguage, contact Littleton High School principal Robert Desaulniers [email protected]. K

JUNE2005 735

COPYRIGHT INFORMATION

TITLE: It’s Time to Rethink Teacher Supervision and EvaluationSOURCE: Phi Delta Kappan 86 no10 Je 2005

WN: 0515201226003

The magazine publisher is the copyright holder of this article and itis reproduced with permission. Further reproduction of this article inviolation of the copyright is prohibited.

Copyright 1982-2005 The H.W. Wilson Company. All rights reserved.