To support Texas districts preparing students for the Texas English Language Proficiency Assessment System (TELPAS) speaking assessment, ALL In Learning provides a dedicated question format: Audio Response (TELPAS, graded with AI).
With this question type, students record their spoken responses directly in the Student Portal. Speech-to-text captures their speech in verbatim mode—adding punctuation and capitalization without correcting grammatical errors or phrasing—and submits the transcript into ALL In Learning's AI grading engine using the official TELPAS rubric.
Technical & Browser Requirements
-
Supported Browsers: Google Chrome or Microsoft Edge (desktop or Chromebook).
Note: Speech recognition APIs are not supported in Safari or Mozilla Firefox. Students opening an audio assessment in Safari or Firefox will see an on-screen prompt advising them to switch to Chrome or Edge.
Microphone Access: Students must grant microphone permissions in their browser when prompted.
District Permissions: AI features must be enabled for your campus or district. If your district restricts AI features, this question type will be hidden from the menu.
1. Creating a TELPAS Audio Response Question
Teachers can add this question type when authoring and editing an assessment in the Authoring Tool:
In the Lessons / Assessments tab, open your assessment editor.
Add a new slide/question by choosing Audio Response (TELPAS, graded with AI) from one of the Insert menu.
Enter your speaking prompt or task in the question area.
-
Configure Question Settings:
Attached Rubric: The official TELPAS audio rubric is locked to this question type by default, and points default to 4 (the rubric's maximum score).
Planning time (seconds): Default is
30(allowed range:0–60seconds).Time limit (seconds): Default is
90(allowed range:60–120seconds).Recording attempts: Default is
2(allowed range:1–3attempts).(Values entered outside these ranges will automatically snap back to the allowed bounds).
Click Save Question. The question will appear labeled as
(audio response, graded with AI)in your question list.
2. Activating the Assessment
Because this question type requires microphone input and browser speech recognition, it must be delivered through the Student Portal.
Go to Activate > Post to Student Portal to launch the session.
-
If you attempt to activate this assessment with Clickers, TAPIT, SCANIT, AI Writing Workshop, or AI Math Workshop, the system will notify you:
"This assessment has an Audio Response question, which students answer by speaking in the student portal…"
3. The Student Experience
When students navigate to an Audio Response question in the Student Portal, the interface guides them through the speaking and recording workflow:
-
Microphone Check:
On the first audio question of an assessment session, students see a Test microphone button.
Students speak to confirm their microphone is working. The system confirms what it heard, or alerts the student if no audio was detected.
-
Planning Countdown:
When ready, the student clicks Start.
A planning countdown begins (default: 30 seconds). Students who want to begin recording immediately can click I'm ready to skip the countdown.
-
Recording:
Once recording starts, students see a live "What we hear" caption box and a countdown timer.
The timer turns yellow at 30 seconds remaining and red at 10 seconds remaining.
Recording will automatically stop if the student reaches the time limit, or the student can click Stop when finished.
Natural Pauses: Unlike standard speech-to-text accommodations, this question type will not cut off or stop recording during long mid-sentence pauses (10+ seconds).
-
Reviewing the Response:
Once stopped, the transcript appears in a read-only response box. Students cannot manually type into or edit this box.
Verbatim Transcript: To accurately evaluate language acquisition, speech-to-text preserves original phrasing and grammar errors exactly as spoken (e.g., awkward phrasing or subject-verb disagreement will not be corrected).
A collapsed "What the computer heard, word for word" drawer allows students to view the raw speech engine output.
The status bar displays the total seconds spoken, word count, and current attempt (e.g.,
Attempt 1 of 2).
-
Re-recording & Submitting:
If attempts remain, students can click Record again to wipe the previous attempt and record a new one. (Blank recordings where no sound was detected do not consume an attempt).
If a student reloads or navigates away from the question, their existing transcript and attempt count are preserved.
Students click Submit work for grading. (An answer must be recorded before submission is allowed).
4. Teacher Grading and Reports
Once students submit their responses, teachers can review scores and feedback in the active grading screen:
Viewing Submissions: Click into the student's submission to view the attempt number, duration spoken, word count, and formatted transcript.
AI Scoring & Feedback: The AI evaluates the transcript against the TELPAS speaking rubric and generates both a score and instructional feedback.
Reports: Completed audio responses flow directly into standard ALL In Learning session reports, Student Work Analysis, and progress tracking alongside other AI-graded assignments.
Frequently Asked Questions (FAQ)
Can students type their answers if their microphone isn't working?
No. Audio Response questions disable typing, the Draw tool, and the standard speech-to-text accommodation to ensure the response reflects spoken performance. If a student cannot access a working microphone, have them switch to a Chrome- or Edge-supported device with audio input enabled.
How does this differ from the standard Speech-to-Text accommodation?
Standard speech-to-text accommodations smooth out student grammar and stop listening after short silences. The TELPAS audio response runs in verbatim mode (preserving speech errors without correcting them) and maintains recording through pauses so English learners have time to formulate sentences.