Phoneme Deletion
○In reviewPlatform: Unity · French, Italian · · verified today
scienceThis is a task in the CRAB study. Click to know more.
Description
As part of the CRAB (Computerized Reading-related Assessment Battery), the Phoneme Deletion task assesses phonological awareness by asking participants to listen to a spoken word or pseudoword and respond aloud with the same item minus a target phoneme. Deletion trials are crossed with word/pseudoword lexicality and paired with repetition trials (say the item back unchanged) as a verbal-processing control condition. Spoken responses are recorded via microphone and scored offline through an automatic speech recognition (ASR) alignment pipeline. This task targets phonological manipulation ability, one of the most consistent predictors of reading acquisition in the developmental literature. Authored in Unity for the lab; the catalog preview is a recorded video walkthrough, and the full task is available as a downloadable Unity build.
Contact & Lab
Contributors
Citation
No citation available — contact the task author.
Credits
Developed at the Brain and Learning Lab (University of Geneva) under Angela Pasqualotto (task conception) and Zeno Menestrina (software developer) as part of the CRAB (Computerized Reading-related Assessment Battery) project. Software updated by Hugo Farajallah and Zeno Menestrina (2025). Adapted from Unity to HTML/JS by Hugo Farajallah in 2026.
Domains
Target Population
Languages
Run under
Modalities
Task Outputs
Phonological awareness, via phoneme-deletion accuracy contrasted with a word/pseudoword repetition control condition
tasklexicality- Deletion accuracy (word vs. pseudoword), from ASR alignment
- Repetition accuracy, as a verbal-processing control baseline
| Variable | Role | Type | Description | Unit |
|---|---|---|---|---|
phase | independent | string | Crossed condition presented on the trial Values: word_deletion, word_repetition, pseudo_deletion, pseudo_repetition | - |
config_id | metadata | string | Identifier of the stimulus configuration used for the trial | - |
config | independent | string | The heard word or pseudoword item and target phoneme for the trial | - |
answersstar | dependent | array | ASR-transcribed spoken response | - |
markup | dependent | string | Error markup from ASR alignment against the expected response | - |
measuresstar | dependent | array | ASR-derived per-trial scoring measures from Needleman-Wunsch alignment | - |
relisten_time | metadata | string | Timestamp if the participant requested to hear the item again | - |
file_url | metadata | string | Path to the uploaded audio recording of the response | - |
Provenance
Source: nccr-seed
Fetched: 2026-02-04
Last verified: 2026-07-17 — passed
Other tasks in this suite
CRAB
