The ability to actively reflect on one's own mental processes, or metacognition, plays a significant role in learning and executing complex tasks by interacting with different learning processes. This study examined the influence of metacognition on declarative and reinforcement learning (RL) mechanisms. I collected data from 221 undergraduates (139 female; age from 18 - 22, mean = 19.0) using a within-subjects metacognitive manipulation half way through the stimulus-response (S-R) learning task created by Collins (2018). The metacognitive manipulation aims to activate monitoring by drawing participants' attention to their own learning strategies and task performance. The task consists of learning blocks of differing lengths; the long blocks rely more on RL, while the short blocks can be completed with declarative or working memory processes. Contributions of declarative and RL mechanisms are also assessed through an incidental post-test given after an intervening task. If metacognition differentially affects declarative and RL, we expect a three-way interaction between the task phase (learning/post-test), block type (long/short) and metacognition (before/during). Results showed significant main effects of task phase (F(1,220) = 153.83, p = 6.32e-34), block length (F(1,220) = 527.29, p = 2.79e-102) and metacognition (F(1,220) = 18.00, p = 2.32e-05), with better performance during the learning phase, short blocks, and metacognitive manipulation. A significant phase by metacognition interaction (F(1,220) = 8.21, p = 0.0042) suggested that metacognitive monitoring improved test performance without interfering with learning performance. Future experiments will examine the possible mechanisms by which metacognitive monitoring facilitated long term memory retrieval.

Statistical learning—predictions based on pattern recognition from varied inputs—is critical in language development. People reference previous grammar encounters in their native language to recognize language patterns, and a strong understanding of statistical learning in language could help create interventions that enhance language development and pattern recognition. However, most research studies on statistical language learning have focused on a select group—monolinguals—ignoring the bilingual community that represent 43% of the world’s population. As such, I review current literature on the unique language learning processes of bilinguals and ask the following research questions: Does bilingualism affect language learning speed, and if so, does it increase or decrease it? When subjects first acquired their second language, what key factors (such as age, proficiency, and whether their second language can be considered another native language) are present? Are there advantages or disadvantages associated with bilingualism in regard to statistical learning and language and how does this compare in relation to monolingualism? To answer these questions, I examined 10 articles that focus on languages across numerous learned languages. A preliminary review of the literature shows that bilinguals have different processes of statistical learning in language development compared to that of their monolingual counterparts. They also indicate that age at which a language is acquired further affects these learning processes. Additionally, results could reveal advantages to bilingualism that strengthen these language learning processes. This literature review can inform future research and studies of the effects of more than one language on statistical learning in language development and point to the need for additional research on not only bilinguals, but also trilinguals, etc.

Aptitude tests are commonly used to measure an individual’s potential to learn a new skill or succeed in a new domain. However, psychometric measures of aptitude are constrained by issues of measurement validity and are susceptible to systemic biases. An alternate neuropsychometric approach is to leverage individual differences in task-free brain characteristics to measure aptitude. Previous researchers have employed this approach using task-free electroencephalography (EEG). Specifically, their work has provided evidence that higher power in the beta frequency band (~13-29.5 Hz), particularly over the right hemisphere, predicts as much as 60% of the variance in the rate at which individuals learn a second-language or a programming language. However, it is unclear what drives this observed relation between beta power and learning rate. The current study explores the mechanistic explanation of this relation, and hypothesizes that this relation is driven by either individual differences in motor control or cognitive control more generally. To test these hypotheses, we tested if individual differences in behavioral measures of motor control (i.e., the Stop Signal task) and/or cognitive control (i.e., the Simon task) mediate the relation between task-free beta power and language learning rate. Contrary to our hypotheses, neither measure of cognitive or motor control mediated the relation between beta power and learning rate. The results of the present study extend previous work relating beta power and learning rate, but future work is needed to elucidate the mechanism driving this relation.