Baseline for AAC goals

AAC baselines go wrong in predictable ways. The SLP probes in the therapy room, but the goal is measured at centers. The baseline counts every message, but the goal counts only spontaneous ones. The adult modeled constantly during the baseline and rarely afterward. Each mismatch makes later progress impossible to judge.

This page sets the rules for an AAC baseline, then shows 12 goals with the baseline written into each one, so you can see how the probe and the goal line up. Numbers are illustrations, not benchmarks. For the full goal bank, see the AAC goals examples page and the speech therapy IEP goals pillar.

Rules for an AAC baseline

1. Probe in the goal's routine. If the goal says "during lunch," take the baseline at lunch.
2. Use the goal's partner support. If the goal says "with partner modeling," model during the baseline at the level adults will keep up. Write down how many models you gave.
3. Use the goal's unit. Messages per 30 minutes, opportunities out of 5, seconds to find a word. Pick it before the first probe.
4. Code spontaneous and prompted separately. The AAC data collection page explains the codes.
5. Take at least 3 probes on different days. Report the median and the range.

The general steps are on the how to collect baseline data page. The IEP must say how progress will be measured,^[1] and the baseline is the first point on that measure.

When the baseline is zero

A zero baseline is common in AAC and it is real data. It often means one of three things: the student has had little modeling, the vocabulary they need is not on the system, or the access method is too hard right now. Check each before you write the goal.

Run a short teaching probe. Model the target 10 times across a routine, then count what the student does in the next 10 minutes. If the student uses the target after modeling, the goal can build from there. If nothing changes, look at the vocabulary and the access method first. AAC has no prerequisites, so a zero is never a reason to hold off on a system.^[3]

How to choose what to measure

Requesting and core word baselines

1. During snack with 10 set-up opportunities (items in view but out of reach), Kai will request with his speech-generating device in 8 of 10 opportunities with no more than a gestural prompt, across 3 days. Baseline: median 2 of 10 across 3 probes, range 1 to 3. Template.
2. Across centers and snack with partner modeling, Lila will use at least 4 different core words (go, more, stop, help, not, like) per 30-minute block on 3 of 4 days. Baseline: median 1 different core word ("more") across 3 blocks. Template.
3. During preferred activities, Ben will build a 2-symbol request (such as "want car," "open please") in 7 of 10 opportunities across 3 sessions. Baseline: median 1 of 10, single symbols only. Template.

Weak baseline: "Kai requests sometimes with the device." Measurable baseline: "In 3 snack probes of 10 set-up opportunities, Kai requested with the device in 1, 2 and 3 (median 2 of 10), with gestural prompts."

Combination and grammar baselines

1. During centers with partner modeling, Maya will combine 2 symbols to request or comment at least 6 times per 30-minute block across 3 days. Baseline: median 1 combination per block (range 0 to 2) across 3 blocks, with adults modeling 12 to 15 times per block. Template.
2. During a book or game activity, Jonah will combine 3 words on his device at least 4 times per 20-minute session across 3 sessions. Baseline: median 0 three-word messages and 3 two-word messages per session. Template.
3. When telling about a past event (weekend, field trip), Ruby will use a past tense verb in 4 of 5 messages across 3 sessions. Baseline: 0 of 5 in each of 3 probes; uses the base verb ("go park"). Template.

Weak baseline: "Maya is starting to combine symbols." Measurable baseline: The first goal in this section. It gives the median, the range and the adults' modeling rate, so a later rise can be read against the same support.

Commenting and social baselines

1. During shared reading with partner modeling, Theo will comment on the book with the device (such as "funny," "uh oh," "big dog") at least 3 times per 10-minute book, across 3 books. Baseline: median 0 comments, 4 answers to questions per book. Template.
2. During free play, Nia will start an interaction with a peer using the device at least 2 times per 20-minute period on 3 of 4 days. Baseline: median 0 initiations to peers, 2 to adults. Template.
3. During lunch with a peer partner, Sam will keep a device conversation going for at least 3 turns on 3 of 4 opportunities. Baseline: median 1 turn across 4 probes. Template.

Weak baseline: "Theo doesn't comment much." Measurable baseline: The first goal in this section. It shows comments against answers, which tells the team the student can use the device but only when asked.

Operational and strategic baselines

1. When asked to say a word stored in a folder, Ella will navigate to it in no more than 10 seconds in 8 of 10 probes. Baseline: median 4 of 10 within 10 seconds; median time 22 seconds. Template.
2. When a partner does not understand her message, Zara will repair it by rephrasing, adding a word or switching modes in 3 of 4 opportunities. Baseline: walked away in 3 of 4 across 3 probes. Template.
3. Using eye gaze, Omar will select the named symbol from a field of 8 in 8 of 10 trials across 3 sessions. Baseline: median 4 of 10 from a field of 8; 9 of 10 from a field of 4. Template.

Weak baseline: "Ella is slow on the device." Measurable baseline: The first goal in this section. It reports both accuracy within a time limit and the median time, so either can improve.

How to measure progress after the baseline

Keep measuring the way you took the baseline. Same routine, same unit, same codes. Probe every 2 weeks for frequency goals and weekly for navigation or access goals, which change faster.

• Frequency goals: the frequency data sheet with columns for spontaneous and prompted messages.
• Opportunity goals: the opportunity trial data sheet.
• Navigation and access: accuracy with a time limit, on the accuracy data sheet.

Record adult modeling next to the student's data each time. The aided language modeling page explains how to count it. For students with few or no spoken words, the measuring goals for nonverbal students page adds options.

Adapting the baseline by grade and access

The IEP team must consider whether the student needs assistive technology.^[2] A baseline taken during a device trial helps that decision. Full goal wording is on the AAC goals page and the IEP goals for AAC users hub. Check each draft in the goal checker.

Frequently asked questions

Use it in IEP Atlas

• AAC goals (speech language)
• AAC in the IEP glossary
• IEP goals for AAC users
• Baseline in the IEP glossary
• Frequency data sheet
• IEP goal checker

Related guides

• Speech therapy IEP goals: examples by area
• AAC goals examples
• How to measure AAC use (spontaneous vs prompted)
• Aided language modeling for AAC users
• AAC in the IEP: devices, training and goals

Sources

1. 34 CFR §300.320: Definition of individualized education program, Legal Information Institute, Cornell Law School
2. 34 CFR §300.324: Development, review, and revision of IEP, Legal Information Institute, Cornell Law School
3. Augmentative and Alternative Communication (AAC), ASHA

Written and checked by the IEP Atlas editorial team under our editorial policy. This guide describes general practice and federal rules. It is not legal advice; your state may add requirements, and the IEP team makes the decisions for each student.