Listening Automaticity: A Reduction of Dual-Task Interference and Working Memory Demands

Bazan Rios, Bartolome Jose Temple University ProQuest Dissertations & Theses 2024 해외박사(DDOD)

According to cognitive psychologists, consistent practice (i.e., tasks with a high degree of similarity) of a skill leads to the development of automaticity, with the degree of automatization being increased if the practice also involves exact repetition. Practice is skill specific, meaning that practicing one skill does not automatize related skills. Once achieved, automaticity has been characterized as the ability to perform a primary task with little interference from performing a concurrent secondary task. A second key characteristic of skill automatization is that it involves a gradual decrease in executive working memory (EWM) demands. The primary purpose of this study is to examine whether such claims put forth by cognitive psychologists are transferable to the automatization of second language (L2) bottom-up listening skills. The secondary purpose of this study is to explore whether reading while listening promotes listening automaticity to a greater extent than listening only, as has been argued in the L2 literature (Chang, 2011; Chang & Millett, 2014; Chang and Millett, 2016; Chang et al., 2018). The third purpose of this study is to examine whether visuospatial working memory (VWM) is a predictor of listening automaticity growth and to what degree it is related to EWM.Although L2 listening is regarded as a cognitive skill, the automatization of L2 listening has not been researched as described by cognitive psychologists. That is, using dual tasks and a reduction in EWM involvement as indices of automaticity, and providing consistent practice with an exact repetition component. In addition, the existing research into the question surrounding what kind of practice is more effective in bringing about L2 listening automaticity -listening while reading or listening only¬- has methodological flaws that have led to an inconclusive answer. To fill these gaps in the literature, I designed a listening-while-reading and a listening-only intervention that blended elements claimed to be conducive to automaticity development in cognitive psychology and L2 research, namely consistent training and exact repetition of stimuli and time pressure, while taking into account the methodological issues of previous L2 research. Data were collected from 290 Japanese high school students, who were assigned to the listening-while-reading, the listening-only, or a control group. Participants in the listening-only intervention listened to texts composed of a narrow range of vocabulary and grammar structures, which warranted consistency, whose speech rates increased by 10 words per minute (wpm) over the sessions from 100 wpm to 180 wpm. The listening-while-reading intervention involved listening to the same texts while reading the transcripts. Exact repetition of the stimuli was provided in both interventions by playing the texts two consecutive times in each intervention session. After the second listening, comprehension was assessed using a Japanese summary of the texts and a set of five multiple-choice questions. Degree of automatization was measured at three time points (i.e., Pretest, Posttest 1, and Posttest 2) separated by two cycles of training, through two dual tasks at each time point. The dual tasks, which consisted of listening to a text delivered at 160 wpm while drawing a series of vertical lines on blank paper and were followed by the abovementioned comprehension measures, were used to construct measures of automaticity. Dual-task data were linked through a Rasch longitudinal design, keeping text difficulty invariant across the dual tasks, with Rasch person measures for comprehension for subsequent analyses computed by combining scores on summaries and multiple-choice questions, while giving double weight to the summaries.Regarding the memory constructs, EWM was assessed through a speaking and a listening span task, whose scores were combined into a global EWM Rasch measure that was used in the analyses to test if the participants' EWM costs diminished over time. Similarly, a single VWM Rasch measure was calculated by combining the participants' scores on the Mr. Peanut task (DeAvila, 1974), which assesses visual memory, and the Corsi block span task (Corsi, 1972), which assesses spatial memory, as additional predictors of L2 listening automaticity. To my knowledge, measures of VWM have not previously been employed as predictors of listening ability.Results from a repeated-measures ANOVA and a multi-group latent growth model indicated that both intervention groups automatized their listening skills to a greater extent than the control group with the listening-only group developing automaticity to the largest extent. Although a reduction of EWM demands was not observed, possibly because the participants were not complete beginners and might have started the study with some degree of automaticity, the fact that they could perform the dual tasks less effortfully over time evidenced automatization. Interestingly, VWM significantly predicted growth in listening automaticity, which was interpreted to mean that as listeners automatize their listening skills, they become able to see what they hear. This result was explained in terms of attentional capacity limits, where higher-level processes, such as visualizing input, cannot be achieved until the lower-level processes such as word recognition or grammatical parsing have been automatized. Follow-up qualitative interviews corroborated the quantitative findings and extended them by suggesting that a small number of training sessions might be sufficient to develop listening automaticity. Results of an additional correlation analysis further showed that VWM and EWM are not highly related as has been shown by some experimental evidence in the working memory literature.The finding that listening only was superior to listening while reading is important because it contradicts previous research and suggests that to automatize listening, one needs to practice in real-operating conditions. Practical and theoretical implications are discussed. First, listening to texts whose speech rates increase over the practice sessions promotes listening automaticity. Next, claims by cognitive psychologists regarding the development of skill automaticity are largely generalizable to the natural conditions of classroom-based research focusing on the acquisition of real-world skills such as L2 listening. Last, the multi-component versus the unitary view of working memory.

작업기억 용량과 컴퓨터 기반 텍스트 양식이 텍스트 이해에 미치는 효과

김민선 연세대학교 대학원 2004 국내석사

Visuospatial working memory assessment using a digital tablet in adolescents with attention deficit hyperactivity disorder

현기정 중앙대학교 대학원 2017 국내박사

Objective: Attention-deficit hyperactivity disorder(ADHD) is a neurodevelopmental disorder hypothesized to have impaired visuospatial working memory (VSWM). However, there are few studies utilizing neuropsychological tests measuring VSWM in ADHD adolescents. The Rey-Osterrieth complex figure test (ROCF) is commonly used as a neuropsychological test that assesses visuospatial working memory for individuals with ADHD. We assessed working memory using the ROCF test in the digital Galaxy tab with the technically new Gaussian filter method. Methods: Thirty adolescents with ADHD and 30 healthy control adolescents were recruited for participation in the current study. All adolescents were assessed with K-WISC-IV, Children’s depression inventory, Korean ADHD rating scale. All adolescents were asked to copy the ROCF using a wireless pen on a Galaxy tab screen while seeing the original paper ROCF figure. Results: There was a significant difference in representative value of the deviation of the original images from template images (R-value) at copy and delayed recall between ADHD adolescents and healthy adolescents. There was no significant difference in R-value of immediate recall between ADHD adolescents and healthy adolescents. In all adolescents (ADHD and healthy) and ADHD adolescents, the R-value of copy was negatively correlated with visuospatial index and working memory index. And in all adolescents and ADHD adolescents, the R-value of delayed recall was negatively correlated with WMI. The R-value of copy and delayed recall was positively correlated with K-ARS in all adolescents and ADHD adolescents. Conclusion: ADHD adolescents showed the differences in the R-value of copy and delayed recall in the digital ROCF version, compared to healthy adolescents. The digital ROCF assessment tool can represent different patterns of visuospatial working memory abilities in ADHD adolescents, compared to healthy adolescents. 목적: 주의력결핍과잉행동장애(ADHD)는 시공간 작업기억력의 손상을 보이는 신경발달학적 장애이다. 하지만, ADHD 청소년에서 시공간 작업기억력을 측정한 신경심리학적 검사에 관한 연구는 많지 않다. 레이-오스테리스 복합도형 검사(ROCF)는 시공간 작업기억력을 평가하는 신경심리학적 평가도구이다. 이 연구에서 갤럭시 탭을 이용하여 ROCF 검사를 한 후 새로운 Gaussian 필터 기법을 통해 작업 기억력을 평가하였다. 방법: 30명의 ADHD 청소년과 30명의 정상 대조군이 이 연구에 모집되었다. 모든 평가군은 K-WISC-IV와 아동 우울 척도, 한국형 ADHD 평가 척도(K-ARS)를 통해 평가 되었다. 모든 청소년은 종이로 된 ROCF를 보고 갤럭시 탭의 스크린에 갤럭시 펜을 이용하여 ROCF를 그리도록 하였다. 결과: ADHD 청소년과 정상 대조군 사이에 복사와 지연 회상에서 유의미한 원본 그림과 Template 그림 간의 차이인 대표 값 ( R-value) 의 차이를 보였다. 그러나 ADHD청소년과 정상 대조군 사이에 즉각 회상에서 유의미한 R-value 의 차이를 보이지 않았다. 모든 청소년과 ADHD 청소년에서, 복사의 R-value 는 시공간 척도와 작업 기억력 척도와 음의 상관관계를 보였다. 그리고 모든 청소년과 ADHD 청소년에서, 지연 회상의 R-value 는 작업 기억력 척도에서 음의 상관관계를 보였다. 복사와 지연회상의 R-value 는 모든 청소년과 ADHD 청소년에서 K-ARS와 양의 상관관계를 보였다. 결론: ADHD 청소년은 디지털 ROCF 검사에서 복사와 지연 회상에서 정상대조군과 비교했을 때 유의미한 R-value의 차이를 보였다. 디지털 ROFC 검사 도구는 정상대조군과 비교했을 때, ADHD 청소년에서 시공간 작업 기억 능력의 다른 양식을 보이는 것을 제시한다.

(The) reducing effect of visuospatial cognitive load on alcohol craving in problematic drinkers

박미영 중앙대학교 대학원 2020 국내석사

본 연구는 문제적 음주자의 시공간 작업 기억 고갈이 알코올에 대한 침습적 사고의 정교화 과정을 방해하여 알코올 갈망을 감소시킬 수 있는지를 확인하였다. 음주 패턴에 문제가 있다고 보고한 90명의 성인을 대상으로 알코올 사용장애 진단검사를 실시하여 50명의 문제적 음주자를 선별하고 시공간적 인지 부하 집단(Corsi 블록 두드리기 과제)과 비인지 부하 집단(정지화면 단순 응시 과제)에 각각 절반씩 할당하였다. 모든 피험자에게 알코올 갈망(알코올 충동 척도 설문지)과 알코올 주의편향(아이트래커를통한 음주와 비음주 사진에 대한 자유 응시 과제)을 집단과제 전후에 반복 측정하였는데, 자유 응시 과제에서는 음주 사진에 대한 초기 초점 경과 시간과 총 초점 응시 시간을 측정하였다. 실험 결과, 시공간 인지부하 집단은 인지부하 과제 이후 알코올 갈망은 감소하였고 자유 응시 과제에서 음주 사진에 대한 총 초점 응시 시간은 과제 후에도 변화가 없었으나, 비인지 부하 집단은 오히려 초점 응시 시간이 증가하였다. 즉, 시공간 인지부하를 경험하게 되면 음주갈망이 줄어들었다고 보고하였으며 비인지부하에서 보이는 음주사진에 대한 주의편향은 나타나지 않았다. 본 연구에서는 시공간 인지 부하를 경험하면 알코올에 대한 갈망이 증가하지 않는다는 것을 밝혔고, 이러한 갈망의 변화를 주의 편향을 통해 확인하였다. This study confirmed that the depletion of visuospatial working memory of problematic drinkers could hinder the elaboration process of intrusive thoughts about alcohol, thereby reducing the alcohol craving. An online survey with Alcohol Use Disorder Identification Test questionnaire was conducted on 90 adults who reported problems with drinking patterns, screening 50 problematic drinkers. They were assigned into Load Group (conduct The Corsi block-tapping task) and No Load Group (gaze still picture in laptop computer screen) in half. Alcohol craving (Alcohol Urge Scale questionnaire, alcohol VAS questionnaires) and attentional bias to alcohol (free-viewing task with alcohol and non-alcohol stimuli through eye-tracker) were repeatedly measured before and after the task, and time to first fixation and total fixation duration to alcohol stimulus were measured when they performed free-viewing task. The result showed that the Load Group has reduced their alcohol craving after visuospatial cognitive load task, and that the total fixation duration on drinking picture in free-viewing task had not changed after the task, whereas the No Load Group has rather increased their total fixation duration. In other words, experiencing a visuospatial cognitive load has reduced the alcohol craving, and there has been no attentional bias toward drinking picture which is shown in No Load Group. This study revealed that when the problematic drinkers experience visuospatial cognitive load, the alcohol craving does not increase, and this was identified through attentional bias to alcohol.

지적장애 아동 대상 시각 작업기억 과제 난이도별 강화 효과 분석 : 분산 강화와 누적 강화의 비교

진동주 연세대학교 대학원 2020 국내석사

The effectiveness of accumulated reinforcement on participants with developmental disabilities has been emphasized over distributed reinforcement, but the generalized effect of accumulated reinforcement is hampered in part by lack of controlling task characteristics or participant characteristics. The present study evaluated if the efficacy of, and preference in children with intellectual disabilities who were 7 years old or older and under 13 years old for distributed and accumulated reinforcement were different between two task difficulties of a visuospatial working memory task. In this study, 77 children with intellectual disabilities conducted four sets of visuospatial working memory(distributed and accumulated reinforcement in an easy task, distributed and accumulated reinforcement in a difficult task). The performance was evaluated by accuracy rate, responses per minute, and correct responses per minute. preference was evaluated by three-point likert scale and selection between distributed and accumulated reinforcement. The increase of responses per minute and correct responses per minute in the accumulated reinforcement was greater than the distributed reinforcement for the difficult task. Although preference for schedules varied across participants, they tended to maintain their selection of reinforcer schedules across task difficulties. Implications and limitations of current research and suggestions for future research were discussed. 발달장애를 대상으로 강화의 분배 방법에 따른 훈련의 효과를 조사한 선행연구는 강화가 작은 단위로 분산된 분산 강화보다 한꺼번에 제공되는 누적 강화가 더 효과적임을 보고하지만, 과제 특성과 참가자 특성을 통제하지 못했다는 점에서 제한된다. 본 연구는 만 7세 이상 13세 미만의 지적장애 및 지적장애를 동반한 신경발달장애를 가진 아동을 대상으로 분산과 누적 강화의 효과가 시각 작업기억 과제의 2개 난이도 수준에 따라 달라지는지 조사하였다. 이를 위해, 선정 기준을 충족한 77명의 지적장애 아동을 대상으로 네 개 조건(쉬운 과제에서의 분산 강화와 누적 강화, 어려운 과제에서의 분산 강화와 누적 강화)에서의 시각 작업기억 과제를 실시하였고, 그 효과를 평가하기 위해 수행(정답률, 분당 응답수, 분당 정답수)과 선호도(3점 척도 자기보고, 강화 조건 선택 비율)를 분석하였다. 연구 결과, 수행의 경우 쉬운 과제에서는 강화 조건에 따른 차이가 나타나지 않았으며, 어려운 과제에서는 누적 강화 조건에서의 분당 응답수 및 분당 정답수가 높은 것으로 나타났다. 선호도의 경우, 네 개 조건에서의 3점 척도 자기보고 및 쉬운 과제와 어려운 과제에서의 강화 조건 선택 비율의 차이가 유의미하지 않았다. 이와 같은 결과의 의의와 함께 본 연구의 한계점 및 후속 연구의 방향을 논의하였다.

Development of an explainable machine learning-based multichannel EEG analysis system for cognitive impairment assessment in sleep-related brain disorders

김현 Graduate School, Yonsei University 2024 국내박사

Machine learning technology, especially deep learning, which has rapidly developed mainly in image and language processing, has been widely applied to medical data analysis. Recently, diagnostic systems that recognize patients with high accuracy have been developed based on various types of medical images, such as radiation and ultrasound imaging. Medical imaging, such as computed tomography and magnetic resonance imaging (MRI), is widely used to diagnose brain disorders. These imaging systems capture the structure of the brain to find structural brain abnormalities. On the other hand, an electroencephalogram (EEG) and functional MRI are mainly used to measure the functional activity of the brain. An EEG is a time-series signal that records the electrical activity of the cerebral cortex using electrodes attached to the scalp. EEGs are mainly used in the field of cognitive neuroscience because they are effective in tracking rapid temporal changes in brain activity when performing cognitive tasks. However, determining which cerebral area exhibits abnormalities is difficult due to the low spatial resolution of an EEG. This study, therefore, aimed to develop an explainable machine learning-based EEG analysis system. The system estimated cortical activity from an EEG through source localization to resolve poor spatial resolution and generate images. Then, the system was designed to diagnose and examine patients with sleep-related brain disorders based on these images. The developed system was applied to experimental data for verification. Up to 99% accuracy was achieved in distinguishing patients from normal controls using multichannel EEG recorded during a cognitive task. The abnormal brain activity characteristics of patients were closely related to cognitive impairment. In addition, the developed system classified most brain activity in patients who received medication as normal. The degree of normalization of brain activity was associated with improved symptoms and sleep quality in the patients. The developed system could be used for diagnosis and treatment monitoring for patients with sleep-related brain disorders. Furthermore, the system might be able to predict treatment response or the possibility of transition in neurodegenerative diseases. 이미지, 언어처리 등을 중심으로 급격히 발전한 기계학습 기술, 특히 딥러닝은 의료 데이터 분석에도 널리 적용되었다. 최근에는 방사선, 초음파 영상 등 각종 의료영상을 이용하여 높은 정확도로 환자를 진단하는 시스템들이 개발되고 있다. 뇌질환 진단을 위해서는 CT, MRI 등의 의료 영상이 많이 쓰인다. 이러한 영상들은 뇌의 구조를 촬영하여 정상인과 차이를 보이는 환자의 구조적 뇌 이상을 판별한다. 뇌의 기능적 활동을 측정하기 위해서는 뇌파, fMRI 등이 주로 쓰인다. 뇌파는 두피에 부착된 전극에서 대뇌피질의 전기적 활동을 기록하는 시계열 신호이다. 인지과제 수행 시 빠른 뇌활동 변화를 추적하는데 효과적이여서 인지기능 연구에 뇌파가 주로 쓰인다. 그러나, 뇌파는 공간적해상도가 낮아 어떤 대뇌영역에서 뇌활동 이상을 보이는지 판단하기 어렵다. 이에 본 연구에서는 설명가능한 기계학습에 기반한 뇌파 분석 시스템을 개발하고자 하였다. 본 시스템은 낮은 공간적해상도를 해소하는 신호원 국소화를 통해 뇌파로부터 대뇌피질활동을 추정하여 영상으로 가공한다. 가공한 영상을 이용하여 본 시스템은 수면관련뇌질환 환자를 진단 및 조사하도록 설계되었다. 개발된 시스템은 수면관련뇌질환 환자의 실험데이터에 적용되었다. 인지과제 수행 시 기록된 다채널 뇌파를 사용하여 정상대조군으로부터 환자를 구별하는 데 최대 99%의 정확도를 달성했다. 환자의 비정상적 뇌활동 특성은 인지기능 장애와 밀접한 관련이 있었다. 또한 개발된 시스템은 약물치료를 받은 환자의 뇌활동 대다수를 정상으로 판단하였다. 뇌활동의 정상화 정도는 환자의 개선된 증상 및 수면의 질과 관련이 있었다. 이로써 본 연구는 개발한 시스템이 뇌파를 기반으로 하는 환자 진단 및 치료효과 모니터링에 활용 가능함을 보였다. 나아가 개발한 시스템은 환자의 치료효과 예측 또는 신경퇴행성 질환으로의 발전 가능성 예측 등에도 활용 가능할 것으로 보인다.