|

Weekly Reflection 3: Testing out Magic School Pro features.

kyrahagel27

I am new to Magic School AI and generally cautious of generative AI. This week, I wanted to try out some of the more unique and specific tools offered by Magic School AI. Below, I will document the prompts I submitted, the AI responses, and my personal response to the Magic School IEP generator. I must admit that I am feeling fairly skeptical about using generative AI to make recommendations or plans for children’s learning. From the beginning Magic School tells you to ensure that you look over the plan in collaboration with professionals, but it makes me uneasy thinking of educators using AI tools to generate ideas or recommendations for individualized learning for students.

Below is a screenshot of the prompt and information I provided to create the IEP:

Below is the AI-generated IEP output:

Student IEP Draft – Review closely before implementation
Responsible Use: This IEP is generated to support educators in collaboration with the student, family, and IEP team—centering their voices, professional judgment, and care in every decision.

Present Levels of Performance

[Student Name] is a 3rd-grade learner who demonstrates strong mathematical reasoning and problem-solving skills, often completing math tasks accurately and quickly. They have difficulty sustaining attention during independent and group work, frequently becoming off-task and occasionally disrupting peers. Reading comprehension of longer, multi-paragraph texts is inconsistent; they understand short texts and explicit information but struggle to infer and synthesize across paragraphs. Written expression is a relative weakness: they can explain logical steps verbally but produce limited written output, with difficulty organizing ideas and maintaining conventions. Emotional regulation and social collaboration are areas of need—[Student Name] shows challenges managing frustration and participating cooperatively in group activities.

Student Needs and Impact of Disability

Because of Autism Spectrum Disorder and ADHD, [Student Name] requires structured supports to access grade-level instruction and participate in classroom routines. Difficulties with attention and self-regulation limit their ability to complete independent work and to engage productively in cooperative learning tasks, which can reduce opportunities for peer practice and oral language development. Reading comprehension deficits for longer texts reduce access to content-area learning that relies on sustained reading and inference. Written expression challenges limit demonstration of knowledge in written assignments and assessments. Emotional dysregulation can lead to interruptions or withdrawal that affect learning time and social participation in age-appropriate activities (recess, classroom group tasks, presentations). Strengths in logical thinking and math provide a foundation for differentiated instruction and motivation.

Goals and Objectives

Measurable Goal: By the end of the IEP year, when given grade-level or scaffolded reading passages of 200–400 words, [Student Name] will accurately answer literal and inferential comprehension questions (who/what/when/where and why/how) with 80% accuracy on 4 out of 5 teacher-created probes, as measured across three consecutive data points.

“Objective”

  • Given a 200–400 word passage and a graphic organizer (main idea/supporting details), [Student Name] will identify the main idea and at least three supporting details on 4 out of 5 trials.
  • When answering teacher-guided inferential questions after reading, [Student Name] will cite textual evidence (1–2 sentences) to support answers with 80% accuracy across 4 of 5 opportunities.
  • With chunked readings (paragraph-by-paragraph) and teacher prompts, [Student Name] will summarize each paragraph in one sentence for at least 3 consecutive paragraphs in 4 out of 5 sessions.

Measurable Goal: By the end of the IEP year, [Student Name] will increase written expression by producing a clear, organized multi-paragraph informative or narrative piece (3–4 paragraphs) with a topic sentence, at least two supporting sentences per paragraph, and a concluding sentence in 4 out of 6 writing assignments, as measured by a teacher rubric.

“Objective”

  • Using a planning template and visual supports (e.g., graphic organizer, sentence starters), [Student Name] will produce a one-paragraph draft with a topic sentence and two supporting sentences in 4 of 5 opportunities.
  • Given explicit mini-lessons on paragraph structure and self-edit checklists, [Student Name] will revise a draft to correct at least 3 identified mechanical or organizational errors (spelling, punctuation, paragraph order) in 4 of 5 assignments.
  • When provided keyboarding access and speech-to-text as needed, [Student Name] will independently produce a 3–4 paragraph draft within a structured time period (e.g., two class sessions) in 4 out of 6 attempts.

Measurable Goal: By the end of the IEP year, [Student Name] will demonstrate improved on-task behavior and social collaboration skills by following a self-monitoring plan and using taught regulation strategies to remain on-task and participate cooperatively in small group or whole-class activities, achieving on-task behavior for at least 75% of monitored intervals across three consecutive weeks.

“Objective”

  • When given a short, individualized behavior plan with a visual schedule and check-in/check-out, [Student Name] will independently complete a self-monitoring checklist for attention and participation during a 30–45 minute work period in 4 out of 5 opportunities.
  • During structured small-group activities, with a peer-support or assigned role, [Student Name] will participate in group tasks (e.g., turn-taking, sharing materials, completing assigned role) without verbal or physical disruption in 4 out of 5 observed sessions.
  • When feeling frustrated or dysregulated, [Student Name] will use taught emotional regulation strategies (e.g., deep breaths, 3-minute cool-down, visual calm-down card) to return to the task within 5 minutes in 4 out of 5 incidents, as measured by staff logs.

Accommodations and Modifications:

  • Preferential seating near the teacher and away from high-traffic distractions; visual boundary cues for work space.
  • Break tasks into smaller, timed chunks with checklists and frequent, scheduled breaks (e.g., work for 10–15 minutes, then 3–5 minute break); use a visual timer.
  • Provide reading supports: pre-teach key vocabulary, offer chunked passages, provide graphic organizers, and allow oral or highlighted-text options for comprehension checks.
  • Writing supports: allow use of speech-to-text or keyboarding, provide sentence starters and graphic organizers, permit extended time for written assignments and tests.
  • Behavior and regulation supports: daily check-in/check-out with a trusted staff member, a personalized self-monitoring chart with taught reinforcement, and access to a designated calm-down area when needed.
  • Assessment accommodations: provide breaks during assessments, allow verbal responses or scribing for some tasks, and administer tests in a quiet environment with extended time.

Review this IEP draft with the family and the multidisciplinary team to confirm priorities, add the family’s input, and adjust goals, accommodations, progress-monitoring methods, and services (e.g., speech/language, counseling, special education minutes) as appropriate.

In the above draft, I put in bold “facts” about the students that AI generated without actually being told that information. In my prompt, I wrote that the student has difficulty with reading comprehension, written output, and emotional regulation. AI wrote the below :

They understand short texts and explicit information but struggle to infer and synthesize across paragraphs. Written expression is a relative weakness: they can explain logical steps verbally but produce limited written output, with difficulty organizing ideas and maintaining conventions. Emotional regulation and social collaboration are areas of need—[Student Name] shows challenges managing frustration and participating cooperatively in group activities.

The draft attaches reasoning and examples to my prompt. For instance, I never said that they could explain logical steps verbally, nor did I say they had difficulty organizing ideas and maintaining conventions; AI made assumptions. They also assumed that the student’s challenges with emotional regulation are related to frustration and difficulty cooperating in groups.

I find the learning objectives that they generated to be based in assumptions and relatively arbitrary with the percentages of achievement assigned.

For example, this learning objective: By the end of the IEP year, when given grade-level or scaffolded reading passages of 200–400 words, [Student Name] will accurately answer literal and inferential comprehension questions (who/what/when/where and why/how) with 80% accuracy on 4 out of 5 teacher-created probes, as measured across three consecutive data points.

In the above passage, it is not specified what scaffolded reading passages are, and the 200-400 word count seems arbitrary, given that they do not know the child’s specific challenges. I do not know what “80% accuracy on 4 out of 5 teacher-created probes measured across three consecutive data points” means. The objective may not work for the particular child, and it does not explain what these scaffolded readings or teacher-created probes should look like. I do not know what data points are or what they should look like for this child.

I find the following quote particularly arbitrary:

“When feeling frustrated or dysregulated, [Student Name] will use taught emotional regulation strategies (e.g., deep breaths, 3-minute cool-down, visual calm-down card) to return to the task within 5 minutes in 4 out of 5 incidents, as measured by staff logs.”

If our goal is to help the student meet their learning objectives and give them tools for emotional regulation, it seems arbitrary to me that they must return within 5 minutes 4 out of 5 times. It also seems unnecessary to log how long it takes them to regulate. For one thing, this goal does not take into account the cause or severity of the child’s dysregulation. If a child becomes frustrated and off-task, but they show an effort to regulate and do not distract others, that is different than if a child becomes violent or leaves the premises. I do not think it is reasonable to expect AI to understand the nuance of a classroom and provide a detailed IEP, so I am not surprised at the result. I also have to take into account that I did not adjust my prompt and may have been given better output if I provided more details or specificity. I do acknowledge that the IEP did provide some regulation strategies and general steps to achieving the goal, which could potentially be used as a rough starting place.

The one part of this that I found helpful was the suggestions for differentiation at the end. I do not think the suggestions should be taken as professional advice or considered a diagnostic IEP, but the teacher could try to implement the suggestions to see if they are helpful.

I added this to the prompt :

Please also consider that the child is not violent, but does occasionally go non-verbal when overwhelmed. They require consistent, but gentle reminders and accountability. Structure and consistent routines are helpful for them. When working with teachers one-one, they can explain logically, but struggle in larger groups.

This is the updated draft IEP after the change in prompt:

Student IEP Draft – Review closely before implementation
Responsible Use: This IEP is generated to support educators in collaboration with the student, family, and IEP team—centering their voices, professional judgment, and care in every decision.

Present Levels of Performance

[Student Name] is a 3rd-grade learner who demonstrates strong mathematical reasoning and problem-solving skills and can explain logical processes clearly in one‑to‑one settings. They have difficulty sustaining attention during independent work and in larger group situations, often needing consistent, gentle reminders to stay engaged. When overwhelmed, [Student Name] may become non-verbal temporarily, withdrawing from tasks until they are supported to regulate. Reading comprehension of longer, multi-paragraph texts is uneven; they read and understand shorter passages well but struggle with inference and synthesis across extended texts. Written expression remains a challenge: they can verbalize steps and ideas but have difficulty producing and organizing multi-paragraph written work.

Student Needs and Impact of Disability

Autism Spectrum Disorder and ADHD affect [Student Name] by limiting consistent access to grade-level instruction without supports that address attention, sensory/communication needs, and social participation. Their tendency to become non-verbal when overwhelmed reduces their ability to ask for help or participate in classroom discussions, which can further block access to instruction and peer interaction. Difficulty working in larger groups and regulating emotions limits participation in cooperative projects, whole-class lessons, and oral presentations—reducing opportunities to practice social communication and collaborative problem-solving. Challenges with sustained attention and task initiation impact completion of independent assignments and performance on timed classroom assessments. Strengths in logical thinking and one-on-one communication suggest that scaffolded instruction, consistent routines, and predictable structure will substantially increase engagement and learning.

Goals and Objectives

Measurable Goal: By the end of the IEP year, when presented with grade-level or scaffolded reading passages of 200–400 words, [Student Name] will answer literal and inferential comprehension questions with 80% accuracy on 4 out of 5 teacher-created probes, as measured across three consecutive data points.

“Objective”

  • Given a 200–400 word passage and a graphic organizer, [Student Name] will identify the main idea and at least three supporting details on 4 out of 5 trials.
  • When answering inferential questions after chunked reading with teacher prompts, [Student Name] will cite textual evidence (1–2 sentences) to support answers with 80% accuracy across 4 of 5 opportunities.
  • With pre-teaching of vocabulary and paragraph summarizing cues, [Student Name] will summarize each paragraph in one sentence for at least 3 consecutive paragraphs in 4 out of 5 sessions.

Measurable Goal: By the end of the IEP year, [Student Name] will produce an organized written piece (3–4 paragraphs) with a clear topic sentence, supporting details, and a concluding sentence in 4 out of 6 writing assignments, using available assistive tools and scaffolds.

“Objective”

  • Using a planning template and visual supports, [Student Name] will create a one-paragraph draft with a topic sentence and two supporting sentences in 4 of 5 opportunities.
  • When provided with speech-to-text or scribing support, [Student Name] will produce a 3-paragraph draft across two structured sessions in 4 of 6 attempts.
  • After a targeted mini-lesson and use of a self-edit checklist, [Student Name] will revise a draft to correct at least 3 identified mechanical/organizational errors in 4 of 5 assignments.

Measurable Goal: By the end of the IEP year, [Student Name] will increase on-task behavior and cooperative participation during small-group and whole-class activities, using a self-monitoring plan and taught regulation strategies to be on-task for at least 75% of monitored intervals across three consecutive weeks.

“Objective”

  • With a daily check-in/check-out and a visual schedule, [Student Name] will complete a self-monitoring checklist for attention and task completion during a 30–45 minute instructional block in 4 out of 5 opportunities.
  • During structured small-group activities with assigned roles and predictable routines, [Student Name] will participate cooperatively (e.g., turn-taking, completing assigned role) without disrupting peers in 4 out of 5 observed sessions.
  • When becoming overwhelmed and non-verbal, [Student Name] will access a pre-taught calming strategy (e.g., written/visual cue, time-limited break, calm-down area) and return to task or communicate needs within 5–10 minutes in 4 out of 5 incidents, as recorded by staff.

Accommodations and Modifications:

  • Preferential seating near teacher and away from high-traffic or noisy areas; provide a clearly defined personal workspace (visual boundary) to reduce sensory distractions.
  • Predictable, consistent routines and a daily visual schedule (posted and reviewed at the start of the day); provide advance notice and countdowns for transitions.
  • Break tasks into smaller, time-bound chunks with written/visual checklists and a visual timer; provide frequent brief breaks and opportunities for movement.
  • Communication supports: use visual choice cards, cue cards, or a simple gesture/sign to indicate the need for a break or support when non-verbal; allow nonverbal responses (pointing, cards) during checks for understanding.
  • Provide one-to-one or small-group instruction for tasks requiring greater expressive language (reading comprehension, written expression), and allow explanation in one-to-one format when large-group participation is required for assessment.
  • Assistive technology and writing supports: permit speech-to-text, keyboarding, graphic organizers, sentence starters, and extended time for writing tasks and assessments.
  • Behavior and regulation supports: daily check-in/check-out with a trusted adult, a personalized self-monitoring chart with gentle reminders and positive reinforcement, and access to a calm-down area staffed with a brief regulation plan; staff trained to provide consistent, gentle prompts and accountability.
  • Assessment accommodations: provide tests in a quiet setting, allow verbal responses or scribing for portions of assessments, permit breaks during testing, and provide extended time.

Review this IEP draft with the family and the multidisciplinary team to confirm priorities, add the family’s input, and adjust goals, accommodations, progress-monitoring methods, and services (for example: counseling, occupational therapy, or additional small-group support) as appropriate.

The revised draft is more specific and does take into account the second prompt, but it continues to add in extra information and use arbitrary percentages and fractions to measure the child’s improvement.

Overall, I feel uncomfortable with the use of AI to create individualized learning plans. Perhaps the AI-generated plan could be a starting place for ideas, but professionals who are familiar with the child’s specific needs must be consulted before implementing or trying any of the objectives or strategies.