Phoneme or Syllable: Monitoring Experiment Data
Description
This data comes from a monitoring experiment that explored how listeners shift their attention between phonemes and syllables when the speech signal changes. The study aims to shed light on how listeners prioritize phonemes or syllables as fundamental perceptual units under changing acoustic conditions. Fluctuations in the masking effect serve as an indicator of attention shifts, calculated from the response times (RTs) of first language (L1) and second language (L2) listeners as they listened to pseudo-words in English (Experiment 1) and Mandarin (Experiment 2). Listeners’ attention was deliberately directed to phonemes through two manipulations: an artificial accent as bottom-up information and prior knowledge of the accent as top-down information. Changes in RTs to phoneme and syllable targets were analyzed under these conditions. In Experiment 1 (English), English listeners showed more flexibility in shifting attention than Mandarin listeners, with syllables being more affected than phonemes. Bottom-up cues had a stronger effect than top-down ones. In Experiment 2 (Mandarin), Mandarin listeners showed more stable syllable processing and less flexibility in shifting attention. Overall, syllable processing was more easily disrupted than phoneme processing—especially by distorted bottom-up cues—suggesting that phonemes may be a more reliable unit of perception, particularly in English. The ability to shift attention depends on a listener’s first language and the sound structure of the language being heard.
Files
Steps to reproduce
The "Experiment 1" folder contains data from Experiment 1, which used English pseudowords and involved English native listeners as well as Mandarin native listeners learning English as a second language (L2). The "Experiment 2" folder contains data from Experiment 2, which used Mandarin pseudowords and involved Mandarin native listeners and English native listeners learning Mandarin as L2. In each folder, the offset.csv files include the offset timing for each stimulus in each list. Listener groups are labeled as follows: EE (English listeners in Experiment 1), ME (Mandarin listeners in Experiment 1), MM (Mandarin listeners in Experiment 2), and EM (English listeners in Experiment 2). Each EE.csv, ME.csv, MM.csv, and EM.csv file represents the original raw data for a specific listener group, collected via the Gorilla online experimental platform. The data_analysis.csv files contain the R scripts used for data analysis. Run the code in these files to analyze the data and reproduce the results.