| In order to investigate the possibility of a working memory (WM) advantage in bilinguals, a spatial WM task, Matrix span task, was developed and validated in Study 1. To be maximally sensitive to the specific processing differences that might be found between monolinguals and bilinguals, the task was non-verbal and based on spatial location information. The main requirement was to recall a series of locations in a fixed order, regardless of the random presentation order. The task was pilot tested with monolinguals and bilinguals in Study 2, and results showed that bilingual and monolingual young adults were matched on all the measures except for the Matrix Span task. Bilinguals scored significantly higher when the scores were calculated from the whole task, including the performance for the string length of 5 and 6. These results suggested that although bilingual young adults showed similar spatial memory capacity (forward Corsi Block), bilinguals outperformed monolinguals when the spatial task required higher levels for processing (rearranging the locations in the Matrix span task), confirming the hypothesis of the bilingual advantage in WM.;In Study 3, 49 monolingual and 45 bilingual young adults performed similarly on all measures of general intelligence, attentional control, processing speed, and attentional scope. The overall performance of the digit WM tasks showed no difference between the two groups, confirming their similar memory capacity. Against these similarities, however, there was evidence for bilingual advantages in performing the spatial WM tasks, especially when the task required higher levels of information processing demand (simple forward version), and when the attentional control demand was higher (complex sequencing version). Regression analysis showed that for monolinguals, both complex reaction time and distractor inhibition contributed to the variance in the WM performance, while for bilinguals, no clear pattern was found. Together these three studies showed bilingual advantages in spatial WM tasks with a different set of underlying abilities contributing to performance for each group. The results were discussed further in terms of their implications for the relationship between executive functions and WM. |
