Week 1: Foundations - AI for Evaluators Bootcamp
WEEK 1

Foundations — FERPA, Tools & Your First Prompts

Set up your evaluation workflow: Choose your AI tool, understand de-identification standards, customize your instructions, and test your first two prompts on a real case.

What You'll Learn This Week

  • FERPA compliance requirements and how they tie to tool choice
  • Green/Yellow/Red rule: Which data types are safe for AI, and which tools handle them differently
  • Your snip-talk-drop workflow across ChatGPT Plus, Copilot, Gemini, and Claude
  • How to build custom instructions that reflect your evaluation philosophy
  • Step 1 (context-setting) and Step 2 (reason for referral) prompts ready to use
  • How to organize your prompt library by FIE section for next-year efficiency

FERPA Compliance & the Green/Yellow/Red Rule

FERPA (Family Educational Rights and Privacy Act) protects student data. The Green/Yellow/Red rule helps you decide what's safe to share with AI, and which tools to use.

GREEN

Synthesized findings, patterns, clinical analysis. Safe to share with AI. Examples: "Weak decoding across English subtests," "Pattern of difficulty with math reasoning."

YELLOW

Assessment data without identifying details. Use with caution. Examples: Raw scores with percentiles (no name, grade, school). Requires careful de-identification.

RED

Student names, family details, health conditions, dates. NEVER share with any AI tool. Never paste these into any system.

The Green/Yellow/Red Tool Rule

Different tools have different data retention policies. This affects what you can safely share and how carefully you need to de-identify.

GREEN TOOLS: School-Provided, Institutional Responsibility

Your district has negotiated data agreements with these vendors. Use these as your primary tools. Your school's legal team has reviewed their privacy policies.

Copilot

Microsoft ecosystem. Use if your district provides Microsoft 365 accounts. Your district has agreements in place.

Gemini

Google ecosystem. Use if your district provides Google Workspace. Your district has agreements in place.

YELLOW TOOLS: Personal, You Manage FERPA Responsibility

You have a personal account. You are responsible for FERPA compliance. Use only the paid versions (data not retained for training). De-identification standards are higher than GREEN tools.

ChatGPT Plus

Paid subscription. Data not used for training. Your personal account, your FERPA responsibility. De-identify carefully (higher standard than GREEN).

Claude

Paid subscription. Your personal account, your FERPA responsibility. De-identify carefully (higher standard than GREEN).

RED TOOLS: Never for Student Data

Free ChatGPT: Data retained for training. Never use for student information, even de-identified.

Your Snip-Talk-Drop Workflow

How you'll actually get information into your AI tool. No typing required.

1

SNIP

Screenshot referral documents, teacher input forms, or observation notes. Paste the image directly into your AI tool. The AI will read and organize it.

2

TALK

Read information aloud from referral documents or your testing observations (including names, dates, and specific details) when you can't safely type or paste it. The AI transcribes and processes it in the conversation. There's no written record created. When you're done, the conversation can be deleted. This is often safer than typing or pasting identifiable data because there's no copy-paste trail or accidental file sharing.

3

DROP

Copy de-identified text from Google Forms, a Word doc, or your notes. Paste it into the AI thread. Remove names and identifying details first.

De-identification checklist: Remove student names, grades, school name, addresses, specific dates. Keep: age, languages, grade level, referral concern, assessment patterns.

Your Custom Instructions: The Foundation

What Are Custom Instructions?

Custom instructions tell your AI tool how YOU work, what frameworks you use, what your district requires, and how to think clinically about evaluation data. Instead of starting from scratch with every prompt, you load these instructions once. The AI then understands your role, philosophy, and standards for all subsequent work.

Example: Instead of telling ChatGPT "I'm a bilingual educational diagnostician" in every prompt, your custom instructions say it upfront. Every response is already tuned to your expertise.

  • You customize the template below for YOUR district framework (CSEP? PSW?)
  • You customize it for YOUR community (rural? urban? bilingual-heavy?)
  • You load it into your chosen tool (Copilot, Gemini, ChatGPT Plus, Claude)
  • Once loaded, every prompt you send benefits from these instructions

SLD/Dyslexia Custom Instructions Template

This is an example template. Download it, customize it for your district and evaluation approach, then load it into your AI tool.

Download SLD/Dyslexia Template (PDF)

Or continue reading to see the template in-page below.

How to Load Custom Instructions

ChatGPT Plus

1. Open ChatGPT. 2. Click "Projects" (left sidebar). 3. Create a new project or select an existing one. 4. Scroll to find the custom instructions section. 5. Paste your customized template. 6. Save.

Copilot

1. Open Copilot. 2. Click "New Agent." 3. Select "Skip to Configure." 4. Click "+" and add "SLD Dyslexia Instructions" (Copilot will draft instructions for you). 5. Go to "Knowledge" and add https://spedsupport.tea.texas.gov/. 6. Click "Create."

Gemini

1. Open Gemini. 2. Click "New Gem" (top left). 3. Give your Gem a name (e.g., "SLD Evaluator"). 4. In the "Instructions" section, paste your customized template. 5. Click "Save."

Claude

1. Open Claude (claude.ai). 2. Click your profile icon (top right). 3. Select "Customize Claude." 4. Click "Create new skills." 5. Paste your customized template as a new skill. 6. Save.

Step 1: Context-Setting Prompt

Send this prompt first, before anything else. It primes your AI to organize your FIE workflow.

Copy and paste into your AI tool: I am an educational diagnostician preparing a Full Individual Evaluation (FIE). I will provide referral documents, assessment data, classroom observations, and intervention information. Please help me organize this information into clear sections for my FIE report. Each section should reflect the document type and the data provided. Use plain language at an 8th-grade reading level. After each section, ask if there are additional details or documents to include before moving to the next section. Never diagnose or determine eligibility. Flag areas where additional information would strengthen the evaluation.

What happens next: Your AI will confirm it's ready. Then you paste your de-identified referral documents (using snip-talk-drop). The AI organizes them into sections and asks clarifying questions.

Step 2: Reason for Referral Prompt

Use this in the same thread after Step 1. Choose the version that matches your referral type.

Option A: Initial Evaluation — Single Suspected Disability

Copy and paste: Please write a short paragraph explaining the reason for an initial evaluation of xStudent. The evaluation will determine if xStudent has a suspected disability in [INSERT DISABILITY HERE], establish their present levels of performance and educational needs, and determine if they require special education services. Write this at an 8th-grade reading level and ensure it is understandable to parents.

Option B: Initial Evaluation — Multiple Suspected Disabilities

Copy and paste: Write a short paragraph explaining that xStudent was referred for a Full and Individual Evaluation (FIE). The purpose is to determine xStudent's strengths and weaknesses, assess characteristics of suspected disabilities, and identify educational needs and strategies to promote academic progress. The suspected disabilities are [INSERT HERE, e.g., "Specific Learning Disability (SLD) and Other Health Impairment due to ADHD"]. Write at an 8th-grade reading level, but keep it professional.

Option C: Re-Evaluation

Copy and paste: Write a short paragraph explaining that xStudent was referred for a Full and Individual Evaluation (FIE) re-evaluation. The purpose is to determine if xStudent continues to have [INSERT DISABILITY], current present levels of performance, and whether changes to their Individualized Education Program (IEP) are needed. Write at an 8th-grade reading level and ensure it is understandable to parents.

What to do: Paste the version that matches your referral. Read the output. Did it need less than 5 minutes of editing? If yes, save it. If no, try adjusting the prompt slightly and test again.

Your Prompt Library Structure

By Week 4, you'll have built a complete prompt library organized by FIE section. Here's what you're building toward:

My Prompts

FIE - Getting Started
Background & History
Cognitive Assessment
Academic Assessment
Impact Statements
Summaries & Eligibility

This week, you'll save Step 1 and Step 2 under "FIE - Getting Started." By Week 4, you'll have 20+ prompts across all sections, plus export your full library as a PDF for next year.

Week 1 Assignment

Complete these steps by Tuesday. Bring your custom instructions and output to the live demo.

  1. Choose your tool: Will you use Copilot or Gemini (GREEN) or ChatGPT Plus/Claude (YELLOW)? Decide based on your district agreements and personal preference.
  2. Customize the template: Download the SLD/Dyslexia template above. Tailor it for YOUR district framework (CSEP? PSW? Hybrid?), YOUR community (rural? urban? bilingual-heavy?), and YOUR evaluation philosophy.
  3. Load your instructions: Follow the step-by-step guide above. Paste your customized template into your chosen tool's custom instructions settings.
  4. Test with Step 1: Open a new thread in your tool. Paste the Step 1 context-setting prompt. Wait for the AI to confirm it's ready.
  5. Test with Step 2: In the same thread, paste the Step 2 prompt (choose A, B, or C based on your current case). Read the output. Does it need significant editing?
  6. Save to My Prompts: Go to the My Prompts section of this bootcamp. Create two new entries: "Step 1: Context Setting" and "Step 2: Reason for Referral [Your Type]." Paste your prompts (and note any tweaks you made).
  7. Reflect: What worked well with these prompts? What would you change for your next case? Come to Tuesday with 1-2 questions.

What's on Tuesday's Live Demo

  • Rita shows her custom instructions (she uses CSEP for most evals in her district)
  • Rita demonstrates her tool choice and why (GREEN vs YELLOW)
  • Live iteration: She runs Step 1 and Step 2 on an actual case and shows how output changes with her instructions
  • Q&A: Bring your custom instructions, your output, and your questions
  • Preview Week 2: Next, you'll build prompts for Background sections (Language, Health, Sociological)
Week 1 Video
Week 2: Background Sections - AI for Evaluators Bootcamp
WEEK 2

Background Sections: Language, Health, Sociological, Emotional

Build five short, focused prompts that organize background information. Each prompt branches based on student profile (monolingual vs. bilingual, with or without diagnosis). Test on your real case.

What You'll Learn This Week

  • Five SHORT, focused prompts (not long walls of text)
  • How to input different referral formats safely (paper, Google Forms, handwritten)
  • Language Background: Branches by monolingual/bilingual AND program type (ESL, Traditional Bilingual, Dual Language)
  • Health/Medical: Core questions for every student, then deeper branching if diagnosis exists
  • When to flag for school psychologist vs. what you (diagnostician) assess
  • Using custom instructions for students with existing diagnoses

Prompt 1: Language Background (Quick Branch)

This is the FIRST prompt. It determines the student's language profile so you know what data to gather next.

Paste into your AI tool: First, I need to understand this student's language exposure at home. Is the student's home language English only, or does the student hear another language at home (Spanish, Mandarin, etc.)? [PASTE OR READ: Home language(s), who speaks what, percentage of exposure in each language]
If Student is MONOLINGUAL ENGLISH at Home:

Go to Prompt 1b (below)

If Student is EMERGENT BILINGUAL (hears another language at home):

Ask: What type of language program is the student in at school? - ESL (English instruction only) - Traditional Bilingual (some Spanish in K-1st, transitions to English in 2nd) - Early Exit Bilingual (similar to Traditional) - Dual Language (instruction in both languages throughout) Once you know the program type, go to Prompt 1c (below)

Prompt 1b: Language Background (Monolingual English)

Use this if student is monolingual English at home.

Paste into your AI tool (same thread as Prompt 1): The student is monolingual English at home. Let me organize the English language development information: [PASTE OR READ: Home Language Survey (if one was completed, include DATE), teacher observations of receptive language, teacher observations of expressive language, any language concerns noted] Please help me understand: Is English language development typical for this student's age? Any red flags?

Prompt 1c: Language Background (Emergent Bilingual)

Use this if student is bilingual (Spanish and English, or another language plus English).

Paste into your AI tool (same thread as Prompt 1): The student is emergent bilingual in [Spanish and English / other languages]. The school program is [ESL / Traditional Bilingual / Dual Language]. Let me organize the bilingual language development information: [PASTE OR READ: Home Language Survey (DATE required), teacher observations of Spanish receptive and expressive, teacher observations of English receptive and expressive, TELPAS scores (if available), LAS or Pre-LAS scores (if available), language dominance testing results (WJ Oral Language or WMLS, if done)] Please help me understand: Is language development typical for a bilingual student in this program? Any red flags in one language or both?
Key Points for Bilingual Students
  • TELPAS: Required in Texas for EB students. Shows English proficiency level.
  • LAS/Pre-LAS: Alternative or additional language proficiency measure (if available).
  • WJ Oral Language or WMLS: Language dominance testing (if done). Tells you which language is stronger.
  • Program type matters: ESL students learn English at school only. Bilingual students get both languages at school. Development looks different in each program.

Prompt 2: Health/Medical Background (Core Questions for Every Student)

Ask these for EVERY student, regardless of diagnosis status.

Paste into your AI tool: Let me gather the student's basic health and medical background: [PASTE OR READ: Vision screening (with DATE), hearing screening (with DATE), pregnancy and birth history, developmental milestones (on time or delays?), is student under care of a physician, current medications (if any), teacher observations of gross motor skills, teacher observations of fine motor skills, is student served under a 504 plan] Please organize this briefly: Vision and hearing status, birth and development, current health status, motor skills, and any accommodations in place (504).

Prompt 2b: Health/Medical (If Diagnosis Exists)

Only use this if the student has an outside diagnosis (ADHD, Cerebral Palsy, autism, etc.).

Paste into your AI tool (same thread as Prompt 2, if applicable): The student has a diagnosis of [ADHD / Cerebral Palsy / other]. Let me provide deeper information: [PASTE OR READ: Diagnosis and date, symptom onset, current medication and response, comorbidities (anxiety, sleep issues, etc.), how symptoms present across settings (home vs. school), prior evaluations, health impairment form from doctor (if ADHD/CP/OHI)] Please help me understand how this diagnosis may be affecting academic performance and what to monitor during evaluation.
Note on Diagnosis and Roles
  • ADHD, Cerebral Palsy, OHI: You may assess deeper. A health impairment form from doctor is needed.
  • Autism or significant emotional/behavioral concern: School psychologist will be more involved. Your focus is on SLD assessment.
  • Use custom instructions specific to the diagnosis (ADHD instructions, etc.) to guide your evaluation.

Prompt 3: Sociological/Family Context (SHORT)

Gather family and home context. Keep it focused.

Paste into your AI tool: Let me understand the student's family and home context: [PASTE OR READ: Family structure, parent education, parent employment, languages spoken at home, books and reading materials in home, housing stability, school mobility history, major life events or stressors] Please organize: Family structure, home literacy, housing and school stability, any major stressors or changes. Before we conclude anything about learning, have you ruled out poverty stress, frequent school moves, limited home literacy, or housing instability as factors affecting learning?

Prompt 4: Emotional/Behavioral (Diagnostician's Lens Only)

This is brief. You're gathering observations, not diagnosing emotional disability.

Paste into your AI tool: Let me gather behavioral and emotional observations from home and school: [PASTE OR READ: How parent perceives student's behavior and emotions at home, how student responds to discipline at home, teacher observations of behavior and emotions at school] Please summarize: Home behavior and emotional functioning, school behavior and emotional functioning. If you notice significant emotional or behavioral concerns that go beyond typical, I'll flag those for the school psychologist.
Important: Your Role
  • You are gathering observations, not assessing for emotional disability.
  • If significant emotional or behavioral concerns emerge, note and flag for school psychologist.
  • School psychologist will do deeper emotional/behavioral assessment if needed.

Reference: SNIP-TALK-DROP Examples

Before you start using the prompts, review how to safely input different referral formats (paper, Google Forms, handwritten notes). This guide shows real examples from Parent Input, Teacher Input, and Classroom Observation forms.

Open SNIP-TALK-DROP Examples (Google Doc)

This guide shows you exactly how to use TALK (read aloud), SNIP (phone photo), and DROP (copy de-identified text) with realistic referral data.

Custom Instructions for Students with Diagnoses

If your student has an outside diagnosis (ADHD, Cerebral Palsy, autism, OHI), load custom instructions specific to that diagnosis into your AI tool. This helps the AI ask the right questions and understand the implications for your evaluation.

ADHD Custom Instructions (TEA Compliant)

Use these instructions when evaluating a student with a documented or suspected ADHD diagnosis. These follow TEA requirements for OHI (Other Health Impairment) evaluation and include the need for a health impairment form completed by a physician.

Open ADHD Custom Instructions (Google Doc)

Includes: TEA compliance, OHI documentation requirements, exclusionary criteria for your community context, bilingual considerations, academic impact focus.

Advanced: Retrofit Your District's FIE Template

You have to use your district's FIE template format. Instead of copying and pasting AI output and reformatting manually, teach the AI your template format upfront.

The Workflow

  1. Copy your district's FIE template section (e.g., Background, Health, Academic Summary)
  2. Paste it into your AI tool: "Here's my district's template format:"
  3. Paste your Week 2 data (from the prompts above)
  4. Ask: "Format this to match my template exactly. Keep it professional and parent-friendly."
  5. Copy/paste the reformatted output directly into your FIE report
Example

Your district's background section format: "Based on current data, sociological, cultural, and environmental factors do not appear to adversely affect the student's learning and behavior patterns to a significant degree."

Give the AI this format + your sociological data from Prompt 3, and it will write a background section that matches your template exactly.

Why This Matters

  • No manual reformatting or copy-paste editing
  • Output matches your district's template on the first try
  • Works for any FIE section (background, health, cognitive, academic, eligibility)
  • Reusable all year with the same template

Week 2 Assignment

  1. Prompt 1: Determine if your student is monolingual English or bilingual. Note the language program type (if bilingual).
  2. Prompt 1b or 1c: Use the appropriate prompt based on language profile. Gather language data.
  3. Prompt 2: Core health questions (every student).
  4. Prompt 2b: If diagnosis exists, use this prompt for deeper health information.
  5. Prompt 3: Sociological/family context.
  6. Prompt 4: Emotional/behavioral observations (brief).
  7. Save to My Prompts Library: Add rows for Prompts 1-4. Note any tweaks.
  8. Reflect: Which exclusionary factors surprised you? Which prompts felt most useful?

What's on Tuesday's Live Demo

  • Rita demonstrates all five prompts on her real case
  • How the branching (mono vs. bilingual, diagnosis vs. no diagnosis) changed what data she gathered
  • How short, focused prompts make the AI more useful than long ones
  • Q&A: Bring your background sections. Ask about the prompts that didn't fit your case.
  • Preview Week 3: Cognitive and Academic Assessment (where the clinical thinking deepens)
Week 2 Video
Week 3: Cognitive and Academic Assessment - AI for Evaluators Bootcamp
Week 3

Cognitive and Academic Assessment

Interpret test results, analyze progress monitoring data, and connect cognitive profile to classroom learning.

Week 3 Overview

This week focuses on cognitive testing and academic performance data. You'll organize test results, progress monitoring data, and work samples to understand the student's cognitive strengths and weaknesses and how they affect academic learning.

Remember: Test scores alone don't tell the whole story. We interpret scores by looking at classroom performance, attendance, consistency of instruction, and bilingual context.

What You'll Learn This Week

  • Seven focused prompts organizing cognitive, progress monitoring, and academic performance data
  • How to reference battery-specific information using NotebookLM as a second brain
  • How to connect cognitive test results to classroom academic performance
  • How attendance and educational history shape interpretation of academic gaps
  • How RTI/MTSS progress monitoring data shows whether student responded to intervention
  • How to analyze work samples against grade-level TEKS standards
  • How to retrofit cognitive and academic findings into your district's FIE template

Texas Education Agency (TEA) Resources

Use these official TEA resources to verify assessment information and understand RTI/MTSS requirements in your district.

Texas Essential Knowledge and Skills (TEKS)

Resource What It Is
TEA TEKS Overview Main TEKS page with links to all subject standards by grade level
TEKS Guide Interactive resource for understanding each TEKS standard with examples and resources (K-8 ELA and Spanish Language Arts available)
English Language Arts and Reading TEKS Grade-level reading standards with web and PDF versions
STAAR Performance Level Descriptors What proficiency looks like on state assessments (grades 3-11)

Response to Intervention (RTI) and Multi-Tiered Systems of Support (MTSS)

Resource What It Is
TEA MTSS Page Official MTSS framework, parental notification templates, intervention requirements
TIER Initiative TEA-funded resources for implementing MTSS: training materials, progress monitoring tools, intervention guides, modules
Texas SPED Support Resource library with MTSS modules, fact sheets, tools, and training opportunities
How to Use These Resources

For TEKS: Go to TEKS Guide and search your student's grade and subject. Print or screenshot the grade-level standard you're comparing to.

For RTI/MTSS: Check the TIER Initiative and Texas SPED Support for your district's specific intervention data and progress monitoring tools. This shows whether the student responded to intervention before referral for special education.

Your Second Brain: Build Your Own NotebookLM

You'll create a personalized NotebookLM that holds information about the assessment batteries YOUR district uses. This ensures you have accurate, verified information you can trust.

Step 1: Set Up NotebookLM

  1. Go to NotebookLM
  2. Click "Create New Notebook"
  3. Name it: "Cognitive Assessment Batteries - [Your District]"

Step 2: Generate and Verify Information About Your Assessment Batteries

Instead of uploading copyrighted assessment manuals, you'll use AI to generate descriptions of your batteries, then verify them against your own manuals:

  1. Start with the Cognitive Assessment Reference Template provided in bootcamp materials (customizable starting point with verified information)
  2. For each assessment battery your district uses (WISC-V, Bateria IV, WIAT-4, KTEA-3, TAPS Spanish, CTOPP-2, etc.):
    1. Step 2a: Ask an AI tool (ChatGPT, Copilot, Claude): "Tell me what [test name] measures, what the indexes/composites are, what standard score ranges mean (e.g., 85-115 for average), and how to interpret bilingual performance on this test."
    2. Step 2b: Open your assessment manual and verify every detail the AI gave you is correct. Check:
      • Are the standard score ranges correct? (85-115, 70-130, etc.)
      • Are the index/composite names spelled correctly and complete?
      • Does the description of what the test measures match the manual?
      • Is any information outdated or incorrect?
    3. Step 2c: If AI information is correct, copy the verified description. If something is wrong, correct it based on your manual.
    4. Step 2d: Add the verified description to NotebookLM (as text, not as a file)

Step 3: Build Your NotebookLM with Verified Information

Create your NotebookLM by adding verified descriptions of the batteries you use:

  1. Add the Cognitive Assessment Reference Template (or your customized version) as your starting point
  2. Add verified battery descriptions that you've confirmed against your manuals (from Step 2 above)
  3. You can paste text directly into NotebookLM as a document, or create a Google Doc with the information and upload the link
  4. Once added, test your NotebookLM by asking it questions you know the answers to
Example: What to Add to Your NotebookLM

You create a document with verified information like:

WISC-V: Measures overall cognitive ability through five indexes. Standard scores: 90-109 average, 85-89 low average, 110-119 high average. Processing Speed and Working Memory may be lower in bilingual students due to language processing load. When verified against WISC-V manual, this is correct.

BATERIA IV: Spanish-normed battery for bilingual Spanish-English speakers. Standard score ranges: 85-115 average. Includes Oral Language and Phonological Awareness subtests relevant to bilingual development. When verified against Bateria IV manual, this is correct.

By verifying first, you ensure NotebookLM has accurate, trustworthy information.

Step 4: Use NotebookLM Based on Your AI Platform

If Using Gemini:

Gemini can connect your NotebookLM directly as a knowledge source in a Gem (Gemini's custom AI tool).

How:

  1. Create a Gem in Gemini
  2. In the Gem settings, find the "Knowledge" section
  3. Click the dropdown and select "NotebookLM" from the available options
  4. Connect your NotebookLM notebook (the one with your verified cognitive assessment information)
  5. Now when you ask your Gem about assessment batteries, it has direct access to your NotebookLM as a knowledge source
If Using ChatGPT Plus or Copilot:

These tools cannot directly access NotebookLM. You'll keep NotebookLM in a separate tab and copy/paste information into your prompts.

How:

  1. Open ChatGPT Plus (or Copilot) in one tab for writing prompts
  2. Open your NotebookLM in another tab
  3. When you need battery information, ask NotebookLM: "What does [test] measure?" or "How do I interpret [score pattern]?"
  4. Copy NotebookLM's answer
  5. Paste it into ChatGPT Plus/Copilot as context: "Based on this information from my assessment reference [paste from NotebookLM], write a test demand explanation for..."
Example: Setting Up Your NotebookLM

You upload: WISC-V manual + Bateria IV manual + Cognitive Assessment Reference Template

You ask NotebookLM: "According to the WISC-V manual, what are the average standard score ranges?"

NotebookLM answers: "85-115 is average range according to WISC-V norms."

You verify: Check your WISC-V manual. Yes, 85-115 is correct. ✓

You use it: Now you can trust this information when you ask NotebookLM other questions about WISC-V.

Prompt 1: Educational History and Access to Instruction

This is the FIRST prompt. It establishes whether the student has had consistent access to instruction and stable school environment.

Paste into your AI tool: Help me organize this student's educational history and access to instruction. How long has the student been in the current district? How many school transitions (moves between schools)? When did each move happen? Was the student in the same language program (bilingual, ESL, general ed) across transitions? [PASTE OR READ: Years in current district, number of school changes, dates of changes, language program consistency, attendance record, gaps in instruction] Based on this information, does the student have stable access to consistent instruction? How might educational transitions affect their academic performance?
Why This Matters

Academic gaps may reflect lack of opportunity to learn, not disability. A student who moved 4 times in 3 years and missed instruction may not have had adequate time to learn phonics or math concepts. This is exclusionary — not a disability.

Prompt 2: RTI/MTSS Progress Monitoring Data

This prompt organizes tier placement, interventions, and progress monitoring data to show whether the student responded to intervention.

Paste into your AI tool (new chat or same thread): Let me organize the student's RTI/MTSS progress monitoring data. What tier(s) has the student been in? How long in each tier? What interventions were delivered? What progress monitoring data is available (CBM scores, fluency, accuracy, graphs)? [PASTE OR READ: Tier placement history, dates, intervention names and duration, progress monitoring data with dates, benchmark comparisons, whether student moved tiers based on progress] Did the student show progress in intervention? Did they meet benchmarks? Were they moved to a higher tier, lower tier, or stayed in same tier?
What We're Looking For

If student showed NO progress in 2-3 tiers of intervention over time, that's evidence they may need special education. If student showed strong progress and moved out of intervention, special education may not be needed.

Prompt 3: Cognitive Assessment Data

This prompt organizes cognitive test results. How you input scores depends on which AI platform you're using.

How to Input Cognitive Scores (FERPA-Safe Methods)

Using Copilot or Gemini:

These tools can handle images and tables.

  1. SNIP method: Take a screenshot of the score table from Q-interactive or your testing software (make sure no student name, ID, or identifying info is visible)
  2. Paste the image directly into Copilot or Gemini
  3. Ask: "What is the overall cognitive profile here? Areas of strength and weakness? How does this compare to same-age peers?"
Using ChatGPT Plus:

ChatGPT Plus works best with descriptive input, not raw numbers. Ground it with NotebookLM.

  1. Step 1: Open NotebookLM and load the Cognitive Assessment Reference Guide
  2. Step 2: In ChatGPT Plus, describe the scores by composite/subtest name and description only. Example: "Verbal Comprehension: Above average (115). Perceptual Reasoning: High average (108). Working Memory: Low average (82). Processing Speed: Low average (81)."
  3. Step 3: Ask ChatGPT Plus: "Using the Cognitive Assessment Reference Guide in NotebookLM, write a test demand explanation and interpretation of this profile. What does each composite measure? How does this compare to same-age peers? What does the scatter pattern suggest?"
  4. ChatGPT Plus references NotebookLM and writes the interpretation
Always Use NotebookLM for Grounding

No matter which AI platform you use, always have NotebookLM open as your reference. Ask NotebookLM battery-specific questions first: "What does WISC-V Processing Speed measure?" "How do I interpret scatter?" Then use that knowledge in your prompts to the other AI tool.

Bilingual Students

If you gave both Bateria IV (Spanish-normed, bilingual) and WISC-V (English), describe scores in both languages. Ask NotebookLM: "How do I interpret bilingual cognitive profiles?" and include that context in your prompt to Copilot/Gemini/ChatGPT.

Prompt 4: Classroom Academic Performance Data

This prompt organizes all the classroom and progress monitoring data that shows academic performance.

Paste into your AI tool (same thread): Let me organize the classroom academic performance data. [PASTE: MAP scores (reading, math with dates), M-Class/DIBELS benchmarks, iReady diagnostic levels, Amira fluency scores, STAAR testing results, district benchmark performance, grade-level equivalents, teacher ratings of academic skills (reading, math, writing)] What areas of academic strength? What areas of weakness? How is this student performing compared to grade-level expectations and same-age peers?
Organize by Subject

Create three groups: Reading (fluency, accuracy, comprehension), Math (calculation, problem-solving), Writing (mechanics, expression). This helps identify specific academic weakness areas for eligibility determination.

Prompt 5: Parent Input on Learning

This prompt captures parent perspective on the student's academic struggles and strengths.

Paste into your AI tool (same thread): Let me add parent input on the student's learning. [PASTE OR READ: Parent observations about reading, math, writing at home. Does the student struggle with homework in specific subjects? Does parent see same struggles at home that teacher sees at school? How does parent support learning at home? Parent concerns about academic gaps?] Based on parent input, what specific academic areas are they concerned about? Do parent observations match classroom performance data?

Prompt 6: Classroom Observation of Academic Tasks

This prompt captures direct observation of how the student approaches academic learning.

Paste into your AI tool (same thread): Let me add classroom observation during academic tasks. [PASTE OR READ (from observation form or notes): How does student approach reading tasks? Math tasks? Writing? How long to get started? Do they ask for help? What happens when work is difficult? Error patterns? Does student show persistence or give up? How much adult support needed to stay on task?] What did we observe about the student's approach to learning? What supports or accommodations did the teacher use?

Prompt 7: Work Samples Compared to Grade-Level TEKS

This prompt analyzes actual student work against grade-level standards.

Paste into your AI tool (same thread): Let me analyze work samples compared to grade-level standards. [SNIP: Take photos of 2-3 student work samples (reading response, math work, writing sample) and paste the images, OR PASTE: Descriptions of student work with examples] Grade-level TEKS standards for [grade and subject]: [copy grade-level standard the student is supposed to master] How does this student's work compare to grade-level expectations? What skills are missing? What can the student do?
Where to Find TEKS Standards

Go to TEKS Guide or TEA TEKS website and search for your grade level and subject area. You'll find the exact standard the student is supposed to master.

Advanced: Retrofit Your District's Academic Summary Template

After running all seven prompts, you'll have a comprehensive picture of the student's cognitive and academic profile. Now format it to match your district's FIE Academic Summary section.

The Workflow

  1. Copy your district's Academic Summary section format from your FIE template
  2. Paste it into your AI tool: "Here's my district's Academic Summary format:"
  3. In the same message, paste all the findings from Prompts 1-7 above
  4. Ask: "Based on this data, write an Academic Summary that matches my template format. Include cognitive findings, academic performance, RTI/MTSS response, and recommendations."
  5. Copy/paste the reformatted output directly into your FIE report
What Goes in Academic Summary

Cognitive profile + Academic performance + RTI/MTSS data + Work sample analysis + Parent/teacher input = Clear picture of whether student has a disability affecting academics

Week 3 Assignment (Due Before Thursday)

  1. Run all seven prompts on a real case from your current caseload (or use the sample case provided)
  2. Use NotebookLM for battery questions (at least once)
  3. Collect output from prompts 1-7 (save the AI conversation)
  4. Try the Retrofit Template workflow: Copy your district's Academic Summary format, paste it with your data, ask AI to reformat
  5. Save to My Prompts Library: Add rows for Prompts 1-7. Note any tweaks or questions.
  6. Reflect: Did the cognitive profile match classroom performance? What surprised you? Where did RTI/MTSS data change your thinking?

What's on Thursday's Live Demo

  • Rita demonstrates all seven prompts on her real case
  • How she used NotebookLM to understand test batteries
  • How cognitive testing and classroom performance align (or don't align)
  • How RTI/MTSS data informed her thinking about disability vs. lack of intervention response
  • Retrofitting Academic Summary to district template
  • Q&A: Bring your cognitive data, academic performance data, work samples. Ask about specific batteries or student profiles.
  • Preview Week 4: Eligibility and Impact (the final clinical reasoning that brings everything together)
Week 3
Using ChatGPT Plus and Copilot Video
Week 4: Portfolio and Second Brain - AI for Evaluators Bootcamp
Week 4

Portfolio and Second Brain

Organize your complete evaluation system, build templates for every case, and establish a sustainable workflow.

Week 4 Overview

This final week brings everything together. You'll organize your prompt library into reusable categories, build templates for FIE sections and REED documents, create a NotebookLM second brain with curated Texas resources, and establish a system to evolve your library with every new case.

By the end of Week 4, you'll walk away with a complete portfolio ready for immediate use in your evaluations.

What You'll Learn This Week

  • How to use IDEA definitions and TEA criteria to determine eligibility through AI-assisted yes/no questions
  • Four focused eligibility question sets: SLD, OHI (ADHD), Intellectual Disability, and Developmental Delay
  • How to organize your complete prompt library into reusable categories
  • How to build templates for FIE sections, REED documents, and impact statements
  • How to create and curate your NotebookLM second brain with Texas education documents
  • How to establish a system that evolves with every new case

Eligibility Question Sets (Based on IDEA Definitions)

Use these prompts to have the AI answer yes/no questions based on your evaluation data. The AI answers the questions but does NOT make the final eligibility determination — that is your role as the evaluator and the ARD committee's role.

Specific Learning Disability (SLD)
Paste into your AI tool: Based on the evaluation data below, answer YES or NO to these questions (do not make a final determination): 1. Does the student show a significant discrepancy between cognitive ability and academic achievement? 2. Does the student have a processing deficit (working memory, processing speed, phonological processing)? 3. Did the student fail to respond adequately to RTI/MTSS interventions? 4. Have exclusionary factors been ruled out (language difference, sensory issues, lack of instruction, emotional disability)? 5. Does the student show academic weakness in a specific area (reading, writing, math) not attributable to other factors? [PASTE YOUR DATA: cognitive scores, academic performance, RTI/MTSS progress monitoring, classroom observations] Answer each question YES or NO based on the data provided.
Other Health Impairment (OHI) for ADHD
Paste into your AI tool: Based on the evaluation data below, answer YES or NO to these questions (do not make a final determination): 1. Is there documented medical diagnosis of ADHD from a physician? 2. Does the student show persistent ADHD symptoms (inattention, hyperactivity, impulsivity) across multiple settings (home and school)? 3. Do ADHD symptoms adversely affect academic performance and learning? 4. Have exclusionary factors been ruled out (trauma, sleep issues, medication side effects, anxiety, learning disability)? 5. Does the student require special education services to address ADHD-related educational needs? [PASTE YOUR DATA: medical documentation, ADHD symptom data, classroom impact, RTI/MTSS response, cognitive and academic performance] Answer each question YES or NO based on the data provided.
Intellectual Disability (ID)
Paste into your AI tool: Based on the evaluation data below, answer YES or NO to these questions (do not make a final determination): 1. Is the student's overall cognitive ability significantly below average (standard score below 70)? 2. Are adaptive functioning deficits present across multiple domains (self-care, communication, social, home/community)? 3. Are deficits evident in both cognitive ability AND adaptive functioning? 4. Did the condition originate during the developmental period (before age 18)? 5. Does the student require special education services due to significantly limited cognitive and adaptive functioning? [PASTE YOUR DATA: cognitive assessment results, adaptive behavior ratings, developmental history, classroom and home functioning, academic performance] Answer each question YES or NO based on the data provided.
Developmental Delay (Ages 3-9)
Paste into your AI tool: Based on the evaluation data below, answer YES or NO to these questions (do not make a final determination): 1. Is the student between ages 3-9? 2. Is there a delay of 15% or more below age level in one or more developmental areas (cognitive, speech/language, motor, social-emotional, adaptive)? 3. Are there signs of developmental delay across one or more domains? 4. Have exclusionary factors been ruled out (sensory impairments, environmental deprivation)? 5. Would early intervention services benefit the student's development and learning? [PASTE YOUR DATA: developmental screening results, areas of concern, educational history, skills assessment, parent and teacher observations] Answer each question YES or NO based on the data provided.

Portfolio Organization: Your Reusable Prompt Library

Organize all prompts from Weeks 1-4 into categories you'll use repeatedly throughout your career.

Suggested Categories

  1. Background Sections (Week 2) — Language, Health, Sociological, Emotional/Behavioral, plus SNIP-TALK-DROP method
  2. Cognitive & Academic Assessment (Week 3) — Educational History, RTI/MTSS, Cognitive Data, Classroom Academic Data, Parent/Teacher/Observation input, Work Samples
  3. Eligibility Determination (Week 4) — SLD questions, OHI/ADHD questions, ID questions, Developmental Delay questions
  4. Template Building — Prompts to generate FIE sections, REED documents, impact statements formatted for your district
  5. Custom Instructions by Diagnosis — SLD/Dyslexia, ADHD/OHI, and future additions (CP, Autism, etc.)
Your My Prompts Library Google Sheet

You started this in Week 1. By end of Week 4, your Google Sheet should have rows organized by category with: Prompt name, What it asks, Platform notes (Copilot/Gemini/ChatGPT), and Any tweaks you made.

Your NotebookLM Second Brain: Curated Texas Resources

Your NotebookLM from Week 3 held assessment battery information. Now expand it to include curated Texas education documents so your second brain supports the ENTIRE evaluation process.

What to Add to Your NotebookLM

Assessment Batteries (From Week 3)

Verified information about WISC-V, Bateria IV, WIAT-4, KTEA-3, TAPS Spanish, CTOPP-2 (AI-generated and checked against your manuals)

IDEA Definitions & TEA Eligibility Criteria

Add key sections from:

  • IDEA disability definitions (SLD, OHI, ID, DD, Autism, Speech/Language, Emotional Disturbance)
  • Texas Administrative Code (TAC) special education requirements
  • TEA eligibility flowcharts and decision trees
TEKS Standards (Grade-Level)

Add grade-level standards for Reading, Math, and Writing for grades you evaluate most frequently (e.g., 3rd-5th grade standards)

RTI/MTSS Framework & Progress Monitoring

Add your district's intervention guides, progress monitoring tools, and tier advancement/exit criteria

Your District's FIE Template & REED Format

Add your district's FIE report sections and REED format so your NotebookLM knows what format you're writing for

Keep Your NotebookLM Current

When TEA updates eligibility requirements, when your district adopts new assessment tools, or when grade-level standards change — update your NotebookLM. Your second brain should evolve with your work.

Build Templates for Immediate Use

With your prompts organized and your NotebookLM built, now create templates you can use on every case.

Template Ideas

  1. Background Sections Template — Run all 4 Week 2 prompts, then retrofit to your district's FIE Background/History section format
  2. Academic Summary Template — Run all 7 Week 3 prompts, then retrofit to your district's Academic Performance section
  3. Eligibility Analysis Template — Run the SLD/OHI/ID/DD yes/no prompts, then compile answers into your eligibility determination section
  4. Impact Statement Template — Prompt the AI to write parent-friendly impact statements for each area of need
  5. REED Re-evaluation Template — Adapted prompts for students already in special ed needing re-evaluation
Save Templates as Google Docs

Create a Google Drive folder "Evaluation Templates" with template shells for each FIE section and REED format. Copy and fill for every new case.

System to Evolve: Learn from Every Case

Your portfolio isn't static. It grows smarter with every evaluation you complete.

How to Evolve Your System

  1. After each case, note what worked in your prompt library (add "works well for", "tweak", or "skip for" notes)
  2. Add new prompts when you discover a prompt structure that works for a particular pattern (e.g., "this prompt works great for gifted LD students")
  3. Update NotebookLM when you learn something new about an assessment tool or eligibility criterion
  4. Customize for your district as templates prove themselves reliable across cases
  5. Archive completed cases (de-identified FIE reports, your notes, template variations used) as reference for future similar cases

Week 4 Assignment (Final Portfolio)

  1. Organize your My Prompts Library into the 5 categories (Background, Cognitive/Academic, Eligibility, Templates, Custom Instructions)
  2. Build your NotebookLM second brain with assessment batteries + IDEA definitions + TEA criteria + TEKS + RTI/MTSS + your district's template
  3. Run the eligibility questions (SLD, OHI, ID, or Developmental Delay) on a real case and save the yes/no answers
  4. Create one template shell — choose Background Sections, Academic Summary, or Eligibility Analysis and build your template version
  5. Set your evolution system — decide how you'll track which prompts work, where to store case notes, and how often you'll update NotebookLM
  6. Walk away with your portfolio — organized prompt library, populated NotebookLM, working templates, ready for immediate use

What's on Thursday's Live Demo

  • Rita demonstrates all four eligibility question sets on a real case
  • How the AI answers yes/no without making the determination (your clinical judgment stays)
  • How to organize your complete prompt library for real-world use
  • How to build and maintain your NotebookLM second brain
  • How to evolve your system with every new case
  • Q&A: Bring your organized prompts, your NotebookLM, your template ideas. Ask how to customize for your specific district needs.
  • Celebration and closing: You now have a complete AI-powered evaluation system ready for immediate use
Week 4
REEDs with Copilot
REEDs with Gemini
Gemini Notebook