How can AI help Alabama evaluators write teacher observations on the ATOT?
How can AI help Alabama evaluators write teacher observations on the ATOT?
Ask a general-purpose AI tool how Alabama evaluates teachers and you will almost certainly get an answer about the Alabama Quality Teaching Standards and EDUCATEAlabama. It will be fluent, specific, and confident. It will also be describing a system Alabama retired in August 2022.
That is not a small error, and it is not the model's fault exactly. There are ten-plus years of AQTS documents indexed on the internet and about four years of Alabama Teacher Growth Program documents. A tool trained on the internet gives you the internet's answer, and the internet still thinks it is 2019 in Alabama.
So before anything else: here is what Alabama actually runs, why the tool on your screen looks nothing like Danielson, and where AI genuinely helps with the write-up.
What Alabama's teacher evaluation landscape actually is
The Alabama Teacher Growth Program (ATGP) replaced EDUCATEAlabama in August 2022. It has four components: a Self-Assessment against the Alabama Core Teaching Standards, the Alabama Teacher Observation Tool (ATOT), a Professional Learning Plan, and an optional Professional Practice Diagnostic.
Each year, a district superintendent picks one of three options through the AIM portal:
Option 1 — Teachers self-assess on all ten Alabama Core Teaching Standards. All classroom teachers are observed on the ATOT. District leaders use the full LEADAlabama indicator set.
Option 2 — Teachers self-assess on a subset of ACTS the district chooses. All classroom teachers are observed on the ATOT. District leaders use a subset of LEADAlabama.
Option 3 — All classroom teachers are observed on the ATOT. Teachers and leaders otherwise use the district's own developed or adopted observation program, and the district stores and maintains its own teacher data.
Notice the constant. Whichever option your board picks, every classroom teacher is observed on the ATOT. That makes it the one instrument every Alabama evaluator has in common, which is exactly why a tool can be built around it precisely.
School administrators are a different track entirely — they go through the APLDS leadership evaluation system, which the School Principal Leadership and Mentoring Act ties to a stipend.
What happened to the AQTS
The Alabama Quality Teaching Standards were the foundation of EDUCATEAlabama. Five standards, thirty-nine indicators, and a set of continuums that rated practice as Emerging, Applying, Integrating, or Innovating.
The State Board replaced the AQTS with the Alabama Core Teaching Standards (ACTS) — ten standards adapted from the InTASC model core teaching standards, grouped into four categories: The Learner and the Learning, Content, Instructional Practice, and Professional Responsibility. Then it adopted a resolution retiring EDUCATEAlabama and stood up the ATGP in its place.
ALSDE's own FAQ puts it about as plainly as a state agency ever does. Asked whether the ATOT aligns with the Alabama Quality Teaching Standards, the answer is one word: No.
A small honesty note, since this is the sort of thing I would want told to me. ALSDE's own manual gives three different adoption dates for the ACTS — the FAQ says the Board approved them in 2019, the key terms page says 2020, and the body text says 2021. State documents do this. The direction of travel is not in doubt; the exact year is genuinely muddled in the source. If you need the date for a board presentation, get it from ALSDE directly rather than from me.
The AQTS is not gone from the buildings, though. It still lives in educator preparation. It still sits in district handbooks nobody has opened since 2020. And under Option 3, some locally adopted observation instruments were written in the AQTS years and were never rewritten. If the form on your desk says AQTS Standard 2, you are not doing anything wrong — you are on a local instrument. It is just not the state system, and anyone who tells you otherwise has not checked.
The five dimensions and twenty-three indicators
Here is the ATOT, in full.
Culture/Climate — a learning environment created and sustained in which each learner is cared for, is safe, has a sense of belonging, and feels secure to share thoughts (4 indicators):
Fosters an environment that embraces all learners
Treats each learner equitably
Encourages learners to share their opinions without fear of negative comments from their peers
Creates enthusiasm for the learning at hand
Learning — processes established to ensure learners have opportunities to demonstrate an understanding of the content and apply that knowledge and acquisition of skills (6 indicators):
Communicates clear explanations about the activities or tasks
Implements lessons and activities that stimulate learners to use higher-order thinking skills
Delivers lessons that are relatable to the learners or aligned to their interests
Monitors learners' understanding of the content and the acquisition of skills
Adapts instruction and activities to meet individual learners' needs
Provides learners with purposeful feedback about their progress and needs
Essentials — core competencies demonstrated by an effective teacher, such as presence and disposition, pedagogical knowledge, content knowledge, and management of the learning environment (5 indicators):
Delivers or facilitates the lesson with knowledge and confidence
Communicates and upholds high expectations for learners' behaviors to maximize learning and well-being
Facilitates use of resources that support learners' needs
Implements instructional strategies that actively engage learners
Manages the learning time in an efficient and optimal manner
Agency — actions that foster learners' ability to self-direct and self-regulate learning (5 indicators):
Empowers learners to be responsible for the learning at hand
Gives learners choices about the learning activities or tasks
Provides assistance for learners to navigate and monitor their learning progress
Encourages learners to persevere with or seek challenging activities or tasks
Builds learners' growth mindset and self-efficacy
Relationships — actions that foster healthy, positive connections between the teacher and learners (3 indicators):
Promotes respectful and caring interactions toward and between learners
Cultivates learner cooperation, collaboration, and inclusivity
Preserves learners' dignity while attending to their individual needs
Twenty-three indicators across five dimensions, and all five dimensions are weighted equally. That is worth sitting with. Relationships carries three indicators and Learning carries six, but the dimensions count the same. A teacher who is strong on instruction and cold with kids does not get to average that away.
Look at the shape of it, too. Two of the five dimensions — Agency and Relationships — are about what the learners are doing and how they are treated. Danielson has nothing quite like Agency as a standalone domain. Alabama decided that whether students can drive their own learning is worth twenty percent of the instrument, all by itself.
The scale is about reach, not adjectives
This is the single biggest thing to understand about the ATOT, and it is the thing every evaluator coming from a Danielson-derived instrument gets wrong at first.
The four levels are Very Evident (4), Evident (3), Somewhat Evident (2), and Not Evident (1). They sound like a quality ladder. They are not. They are pegged to percentages of learners:
RatingReachVery Evident (4)All / fully — more than 75% of the learnersEvident (3)Most / mostly — 51% to 74% of the learnersSomewhat Evident (2)Some / minimally — about 25% to 50% of the learnersNot Evident (1)Less than 25% of the learners
Read that again with a lesson in mind. A teacher runs a genuinely sophisticated questioning sequence. It is well designed, it is rigorous, and about a third of the class is in it while the rest wait it out. On a Danielson instrument you might be arguing about Proficient versus Distinguished. On the ATOT that is a 2.
This is not the tool being harsh. It is the tool asking a different question. Danielson asks how good the teaching was. The ATOT asks how many kids it reached. Those come apart more often than we like to admit, and the ATOT is built so that you cannot rate the performance without counting the room.
It also means the top of the scale is attainable in a way Distinguished usually is not. Very Evident does not require brilliance. It requires that the practice be fully implemented, fully embedded, and landing on more than three-quarters of the learners. A steady, unglamorous teacher whose routines are airtight and whose kids are all in can score 4s all day. I think that is the right instinct.
One rating, three factors
The ATOT form records one rating per indicator. Twenty-three ratings, that is it.
But getting to each rating means weighing three separate factors, and they can disagree with each other:
Implementation of Practices — the extent to which the teacher carries out the observable action with quality, fidelity, and rigor. The manual's descriptors are telling here: adjusting actions to meet learners' needs, demonstrating deep understanding of best practice, changing course based on learners' cues. Implementation is not "did they do the thing." It is "did they do the thing well, and did they adjust it when the room told them to."
Authenticity of Practices — the extent to which the action is embedded in the classroom environment. Do the practices appear familiar to the learners? Do they appear to occur regularly rather than sporadically? This is the factor with no analogue anywhere else in American teacher evaluation, and it is the most interesting thing on the instrument. Authenticity is asking whether the practice lives here, or whether it showed up because you did. Kids give this away instantly. If the turn-and-talk protocol is real, they slide into it. If it is theater for the observer, they fumble it, look at each other, wait for instructions. You can see it in five seconds and you cannot fake it.
Effect on Learners — the extent to which the practice influences the learners. They respond. They make adjustments. They participate and engage. This is the factor that sets the percentage band.
So a practice can be beautifully implemented, obviously novel to the students, and land on half the room. Three factors, three different answers, one rating to give. The manual says to use professional judgment and training to determine the final item rating, which is an honest way of admitting the tool cannot do that arithmetic for you.
The ATOT is formative. That does not mean it is easy.
Here is where I want to be careful, because it would be easy to oversell.
ALSDE is unambiguous that the ATOT is a formative observation tool only:
An LEA cannot use it as its sole tool for evaluation.
Its results are not designed to factor into employment decisions.
Its data should not be placed in personnel files.
Reporting to the state is aggregate at the school and district level. Individual teachers are not reported.
The Teacher Effectiveness Score — 80% ATOT, 20% student growth — is not part of state report card calculations.
If a vendor tells you their tool produces a defensible evaluation from an ATOT, they have not read the manual. There is no such thing.
And yet. Every classroom teacher in Alabama gets observed on this instrument every year, and ALSDE recommends one per semester. That is twenty-three ratings per observation, plus the Formative Feedback Form asking for specific examples of practices observed across all five dimensions, plus the feedback conversation itself. For a principal with forty teachers, that is not a rounding error in the week. It is the week.
The paperwork does not care whether it is formative. It still has to be written, it still has to be specific, and it still has to be good enough that the teacher across the desk finds it useful rather than generic. That is the job. That is what a tool should shorten.
Where the stakes genuinely sit is your district's own instrument — the Option 3 model, or whatever your board adopted alongside the required ATOT. That is the thing that informs employment decisions. That is the thing that lands in a file. Both write-ups take your evening. Only one of them has teeth. It is worth being clear-eyed about which is which.
Twenty minutes
The ATOT observation window is twenty minutes. Formal, informal, announced, unannounced — the state does not fuss much about that, but the window is the window.
Twenty minutes is not very long, and it constrains what you can honestly rate. You are being asked to judge whether a practice is embedded, on the basis of a fifth of a class period. Some of that you can genuinely see — learner fluency with a routine tells you a great deal. Some of it you cannot, and the honest move is to say so rather than to invent it.
Two small mechanics that matter:
If a teacher is observed more than once, only the most recent ATOT counts toward the Effective Classroom Instruction Score. Scores are not averaged.
The ATOT is not for specialty roles. Not counselors, not library media specialists, not instructional coaches, not administrators. Observation data for non-teachers is removed from the calculation.
And one that catches people: the descriptive actions are not built into the platform. ALSDE says observers need to provide the sample descriptive actions to the teacher themselves. So the teacher sees a rating, and unless you hand them the language behind it, they see a number with nothing attached.
Where AI genuinely helps — and where it doesn't
It helps with the write-up. You watched the lesson. You know what you saw. The gap between knowing and having written it down in language a teacher can use is where the evening goes. Mapping evidence to the right indicator, drafting a best-fit rating, and producing specific examples of practice observed for each of five dimensions — that is drafting work, and drafting is what these tools are for.
It helps with the descriptive-action language. Since the platform will not give the teacher the descriptors, feedback written in the ATOT's own vocabulary means the teacher recognizes what they are reading instead of decoding it.
It helps with consistency. Twenty-three indicators, five evaluators, forty teachers, one building. Drift is not a character flaw, it is arithmetic. A tool that anchors every draft to the same rubric language narrows the spread.
It does not know your room. Whether that practice is embedded or performed is a judgment you make from having been in that classroom before, knowing that teacher, and watching those kids. No model has that.
It does not know what share of the room was in it. The percentage bands are the heart of this instrument, and the only person who can count the room is the person who was standing in it. If your notes do not say how many kids were engaged, no tool can invent it — and one that offers to is making things up.
It cannot give you a rating. It can give you a draft rating, which you accept, change, or throw out. The ATOT's whole design leans on evaluator judgment. That is why it names three factors instead of a formula.
What to write down, by dimension
The tool can only work with what you captured. A few notes that make the write-up dramatically easier:
Culture/Climate — Write down who the teacher spoke to and who they did not. Equity on this instrument is about distribution, and distribution is countable. Note the moment a kid took a risk and what happened next.
Learning — Capture the actual explanation, roughly verbatim if you can. Note how many kids started working without asking a follow-up question — that is your reach number for the first indicator. Note any adaptation and what triggered it.
Essentials — Note transitions and how long they took. Note whether kids got materials without asking. Note whether the teacher's content answers were confident or evasive when a question went off-script.
Agency — Note who was working the problem: the kids or the teacher. Note any real choice, and whether it changed the work or just the packaging. Note wait time.
Relationships — Note the redirect. Every classroom has one, and how a teacher redirects a kid in front of peers tells you more about the dignity indicator than anything else in twenty minutes.
And across all of it, note the count. Most, some, nearly all, about half. Your rating depends on it, and you will not remember by that evening.
Common scoring mistakes in Alabama
Reading the scale as a quality ladder. The most common error by a distance. Very Evident is a reach threshold, not a compliment.
Rating on Implementation alone. A polished practice that reaches nobody is not Evident. The three factors are all supposed to move the number.
Ignoring Authenticity. It is the easiest factor to skip, because rating it means admitting you are judging whether the lesson was staged. But that is what the instrument asks, and the kids' fluency tells you.
Treating a 2 as a bad rating. Somewhat Evident means the practice reached a quarter to half the room. On a first-year teacher, in November, that is often an accurate and unremarkable description of reality.
Using an AQTS instrument and thinking it is the state system. Or the reverse — assuming your local instrument is invalid because it is not the ATOT. Both are wrong. You run both.
Rating a counselor on it. The ATOT is not built for specialty roles and their data is stripped out anyway.
Alabama in context
Most states hand districts a framework and a set of requirements and let each board build an instrument. Alabama did something narrower and, I think, smarter: it mandated one twenty-minute formative observation on one common tool for every teacher in the state, and left the actual evaluation system to the district.
That split is unusual, and it explains most of the confusion. Alabama has a statewide observation tool and local evaluation systems. They are different things, they serve different purposes, and the state tool is deliberately the low-stakes one. Once you see that, the ATOT's design makes sense — it is not trying to be a personnel instrument, which is why it can afford to ask about growth mindset and dignity.
It also means an Alabama principal does two write-ups on two instruments for the same teacher. Which is, from where I sit as a teacher who has been on the receiving end of both, a lot of somebody's evening.
How EvalScribe handles Alabama
EvalScribe is an AI-powered evaluation tool built by a practicing K-12 teacher. For Alabama, both frameworks ship.
The ATOT is built in with all five dimensions and all 23 indicators, each with language across the four levels, the descriptive actions an observer watches for, and the distinctions that separate adjacent ratings — including the percentage bands, because in Alabama the band is the rating. The legacy AQTS ships alongside it for districts still running an EDUCATEAlabama-era local instrument, clearly labeled as legacy rather than passed off as the state system. EvalScribe also ships Danielson, Marzano, Stronge, CEL 5D+, NIET, and Kim Marshall, which covers most of what Alabama districts adopt for their own models.
You observe the way you already do. You capture notes by typing, dictating, or photographing your handwriting — Smart Scan OCR handles the handwriting, including the pedagogical vocabulary that trips up generic OCR. EvalScribe maps your evidence to the relevant indicators, drafts best-fit ratings, and writes feedback in the ATOT's own vocabulary, descriptive actions included. You review, edit anything, and export a clean PDF to carry into the feedback conversation.
Honest scope, because it matters more than the pitch:
EvalScribe drafts the write-up. It does not enter your ATOT into Cognia's myJourney, complete a teacher's ACTS self-assessment, or build a Professional Learning Plan. Those live in the state platform and they stay with you.
EvalScribe does not know which instrument your district adopted for its own evaluation model, and under Option 3 that could be nearly anything. If yours reads differently, email [email protected].
EvalScribe will not tell you what share of the room was engaged. That number comes from your notes, because it comes from your eyes.
And on privacy: observation notes are processed transiently on Microsoft Azure and are not retained on our servers. Evaluation records live locally on your device. Nothing is used to train public AI models. ALSDE's position is that ATOT data should not go in a personnel file at all — so where the record lives is not a small question.
Three evaluations are free. The Alabama page is at evalscribe.com/alabama. If you're leading a rollout across a district rather than a building — which option you're running, whether your local instrument is current, evaluator consistency across 23 indicators, access codes, a Data Processing Agreement — that's at evalscribe.com/district.
FAQ
Does Alabama still use the AQTS? Not as the state system. The Alabama Core Teaching Standards replaced it, and the Alabama Teacher Growth Program replaced EDUCATEAlabama in August 2022. ALSDE says directly that the ATOT does not align to the AQTS. The AQTS still exists in educator preparation and in some local instruments that were never rewritten.
What are the five ATOT dimensions? Culture/Climate (4 indicators), Learning (6), Essentials (5), Agency (5), and Relationships (3). Twenty-three indicators total, all five dimensions weighted equally.
What does Very Evident mean? More than 75% of the learners, with the practice fully implemented and fully embedded in the classroom. It is a reach threshold, not a judgment about teacher sophistication.
Is the ATOT an evaluation? No. It is a formative observation tool. It cannot be a district's sole evaluation instrument, its results are not designed to drive employment decisions, and ALSDE says its data should not be placed in personnel files.
Then why does it take so long? Because it is still 23 ratings and five dimensions of specific examples, on every classroom teacher, every year. Formative and quick are not the same thing.
How often is it required? At least once per school year per classroom teacher. ALSDE recommends one per semester. If a teacher is observed more than once, only the most recent ATOT counts.
How long is the observation? Twenty minutes.
Is the ATOT used for counselors or instructional coaches? No. It is not for specialty roles, and non-teacher observation data is removed from the effectiveness calculation. School administrators go through APLDS; district leaders use LEADAlabama.
What is the difference between the eleot and the ATOT? The eleot focuses on students and the extent to which they are engaged in learning. The ATOT focuses on the teacher and on professional growth. Both are Cognia tools; they are not the same instrument.
Does my district have to use the ATGP? No, but it does have to use the ATOT. Under Option 3 a district runs its own developed or adopted observation program and stores its own data — but every classroom teacher is still observed once a year on the ATOT.
Can AI write my ATOT for me? It can draft it. It cannot watch the lesson, count the room, or decide whether a practice was embedded or performed. Those are yours, and they are the parts that require having been there.
References
Alabama Teacher Growth Program Orientation Manual, Alabama State Department of Education — the ATOT with sample descriptive actions (Appendix C), the ATOT Rating Guide Rubric, the ACTS Self-Assessment Rubric (Appendix A), the Glossary of Terms (Appendix B), and the full Alabama Core Teaching Standards (Appendix F).
State Superintendent Memorandum, Teaching Effectiveness / APLDS / ATGP / LEADAlabama Implementation Procedures 2025-2026 (FY25-2061, June 30, 2025) — the three district options and the annual timeline.
Alabama Core Teaching Standards, Ala. Admin. Code r. 290-3-3-.03.
Teacher Effectiveness Score Business Rules, ALSDE — the ATOT's role in the Effective Classroom Instruction Score.
Alabama Teacher Growth Program, alabamaachieves.org — current program page and resources.
Every citation above should be verified against the current ALSDE publication before you rely on it for a board conversation. State documents get revised, and as noted, ALSDE's own manual disagrees with itself about when the ACTS were adopted.
