# README: ERP Experiment on the Perception of Mandarin /u/ pre-diphthongization

## 1. Project Overview

This MATLAB script implements an auditory **Oddball Paradigm** designed to investigate the neurophysiological basis of vowel evolution. Specifically, it examines how the consonantal place of articulation (Acute vs. Grave) influences the perception of the  diphthongization in Mandarin Chinese.

* **Theory**: Featurally Underspecified Lexicon (FUL) & Perceptual Saliency.
* **Target Components**: Mismatch Negativity (MMN) and Late Discriminative Negativity (LDN).
* **Task**: Passive listening (subjects watch a silent movie).

---

## 2. Technical Requirements

* **Platform**: MATLAB (R2016b or later recommended).
* **Toolbox**: **Psychtoolbox-3 (PTB-3)**.
* **Audio Driver**: ASIO drivers are highly recommended for sub-millisecond trigger precision.

---

## 3. Experimental Parameters

The script is programmed with the following constraints to ensure high-quality ERP data:

* **Stimulus Ratio**: 85% Standard / 15% Deviant.
* **Sequence Constraints**:
* Begins with **10 consecutive Standards** to establish a sensory memory trace.
* A minimum of **2 Standards** between any two Deviants to prevent neural adaptation.


* **Timing**:
* **ISI (Inter-Stimulus Interval)**: Jittered between **350 ms and 650 ms** (uniform distribution) to eliminate rhythmic expectancy.


* **Trigger Codes**:
* Standard = `10`
* Deviant = `20`



---

## 4. Stimuli (Private Access)

**Note on Data Privacy**: The audio stimuli (`.wav` files) used in this study are **not included** in this public repository to protect experimental integrity and intellectual property.

To run this script, users must provide their own stimuli named as follows:

* **Acute Block**: `su.wav` (Standard), `sou.wav` (Deviant)
* **Grave Block**: `ku.wav` (Standard), `kou.wav` (Deviant)

*All files must be RMS-normalized, sampled at the same rate (e.g., 44.1 kHz), and have the initial silence removed.*

---

## 5. Instructions for Use

1. Place the main script and your private `.wav` files in the same directory.
2. Run the script in MATLAB.
3. Enter the **Subject ID** and **Block ID** (1 for Acute, 2 for Grave) when prompted.
4. The experiment will pause; press any key to start once the subject is ready and the EEG recording has commenced.
5. Press **ESC** at any time to abort the session and safely close the audio port.

---

## 6. Data Analysis Note

Data should be processed using standard ERP pipelines (e.g., EEGLAB/ERPLAB). Analysis focuses on the **Difference Wave** (Deviant minus Standard) to verify the predicted **asymmetry** in the Acute group versus the **symmetry** in the Grave group.

---