Qualitative and quantitive assessment of new language perception of bilinguals using electroencephalography(EEG)

• Main Author: Yuvadarshinii, Ilang Kumaran
• Other Authors: Xu Chengjie
• Format: Theses and Dissertations
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering::Bioengineering
• Online Access: http://hdl.handle.net/10356/68449
• Institution: Nanyang Technological University

Artificial Grammar Learning (AGL) is a paradigm of study within cognitive psychology and linguistics that tests the subjects’ implicit learning ability to pursue a new or made-up grammar [1]. This interprets language in terms of concepts and its underlying semantics. The conclusions drawn from these studies explain the mental processes underlying in these studies. Previous researches in AGL have shown that bilinguals have an advantage over monolinguals in various cognitive control tasks such as new grammar learning [2]. Based on these studies, it was hypothesized that the bilinguals are proficient in distinguishing grammar from non-grammar sequences. This project aims to investigate and analyze qualitatively and quantitatively, the efficiency of bilinguals in pursuing a new language grammar and check the validity of this argument in using more than one languages. In order to test the hypothesis, EEG responses from six bilinguals for language based visual-auditory stimuli were taken and analyzed. EEG is a powerful non-invasive tool used to explore human brains [16]. The analysis was based on the Event Related Potentials (ERP) components, their latency and amplitudes obtained for each subject in two conditions – Grammar and Non-Grammar in two languages. The observations showed that the subjects showed two components namely N200 and P300 in their ERPs corresponding to involuntary active processing of visual-auditory stimulus and selective attention and decision making respectively. The means of the data obtained in grammar and non-grammar conditions for both languages for all subjects were compared using T-Test assessment. The findings showed that majority of the response of the subjects for grammar sequences and non-grammar sequences were distinguishable with 95% significance. The allowable probability of false positive discoveries in the statistical comparisons were confined to atmost 2-5% significance.