心理因素与学业表现：机器学习分类预测模型

随着高等教育规模的扩大,学业表现不良逐渐成为一个不容忽视的现象,对学业表现不良的学生进行预测并提早给予干预可降低退学率并减少教育资源的损失。由于导致学业表现不良的因素众多且关系复杂,传统的基于相关分析的研究方法很难建立早期预测模型并进行应用。本研究旨在利用机器学习算法,对数据进行挖掘,并建立学业表现预测模型。研究对653名大一新生的心理健康状况、应对方式、人格、内外控倾向和相关人口统计学信息进行了收集,并在一年后采集了其学业成绩,利用随机森林(RF)、K邻近（KNN）、支持向量机（SVM）、决策树（DT）、朴素贝叶斯（NB）等机器学习算法建立了分类模型。结果显示,随机森林算法在识别学业表现不良学生时有最好的表现,其中准确率95.86%, 召回率91.83%,f1分数为93.80%。特征重要性分析显示,前10个对模型有最高贡献度的特征包括：年龄、性别、是否为独生子、内外控倾向、神经质倾向、积极应对倾向、宜人性倾向、一般症状指数、开放性倾向和焦虑水平。为避免过度拟合问题,本研究在一年后收集的166名新生样本中进行了模型验证,结果显示模型在新数据集上有较好的泛化表现,其中f1分数90.90％,准确率92.60％,召回率89.26％。研究提示基于人口统计学和心理测评信息,机器学习算法有助于及早识别学业表现不良学生并为开展早期干预提供启示。     

A systematic review of the effectiveness of machine learning for predicting psychosocial outcomes in acquired brain injury: Which algorithms are used and why?

Clinicians working in the field of acquired brain injury (ABI, an injury to the brain sustained after birth) are challenged to develop suitable care pathways for an individual client’s needs. Being able to predict psychosocial outcomes after ABI would enable clinicians and service providers to make advance decisions and better tailor care plans. Machine learning (ML, a predictive method from the field of artificial intelligence) is increasingly used for predicting ABI outcomes. This review aimed to examine the efficacy of using ML to make psychosocial predictions in ABI, evaluate the methodological quality of studies, and understand researchers’ rationale for their choice of ML algorithms. Nine studies were reviewed from five databases, predicting a range of psychosocial outcomes from stroke, traumatic brain injury, and concussion. Eleven types of ML were employed with a total of 75 ML models. Every model was evaluated as having high risk of bias, unable to provide adequate evidence for predictive performance due to poor methodological quality. Overall, there was limited rationale for the choice of ML algorithms and poor evaluation of the methodological limitations by study authors. Considerations for overcoming methodological shortcomings are discussed, along with suggestions for assessing the suitability of data and suitability of ML algorithms for different ABI research questions.     

The Natural Selection of Prediction Heuristics: Anchoring and Adjustment versus Representativeness

There are several heuristics which people use in making numerical predictions and these heuristics compete for the determination of prediction output. Some of them (e.g. representativeness) lead to excessively extreme predictions while others (e.g. anchoring and adjustment) lead to regressive (and even over-regressive) predictions. In this paper we study the competition between these two heuristics by varying the representation of predictor and outcome. The results indicate that factors which facilitate reliance on representativeness (e.g. compatibility between predictor and outcome) indeed lead to an increase in extremity, while factors that facilitate reliance on anchoring and adjustment (e.g. increased salience of a potential anchor) lead to a decrease in extremity.     

Engagement and Interaction: A Comparison Between Supported Employment and ATCs

The interaction and engagement patterns of 13 supported employees and 38 regular Adult Training Centre (ATC) attenders were compared. Data were obtained through direct observation within 13 employment sites and a representative sample of ATC organised activities. Results indicated that individual involvement was significantly higher in employment, compared to ATCs, but that the ATC group experienced significantly more social interaction, mainly due to higher levels of task specific dialogue between individuals and the service staff within ATCs. A breakdown of with whom interactions occurred revealed that individuals in ATCs tended to interact more often with care staff than anyone else. Individuals tended to interact more often during work with their non-disabled co-workers and a substantial proportion of time was spent interacting with the public. These findings suggest that the supported employment option yields advantages for individuals in some key aspects of community integration, compared to ATCs.     

ON THE BENEFITS OF PHILOSOPHY AS A WAY OF LIFE IN A GENERAL INTRODUCTORY COURSE

Philosophy as a way of life (PWOL) places investigations of value, meaning, and the good life at the center of philosophical investigation, especially of one’s own life. This essay argues that PWOL is compatible with general introductory philosophy courses, further arguing that PWOL-based general introductions have several philosophical and pedagogical benefits. These include the ease with which high-impact practices, situated skill development, and students’ ability to “think like a disciplinarian” may be incorporated into such courses, relative to more traditional introductory courses, as well as the demonstration of philosophy’s value to students by explicitly tying philosophical investigation to students’ own lives.     

Emotional facial expressions affect visual rule learning in 7- to 8-month-old infants

Rule learning (RL) is an implicit learning mechanism that allows infants to detect and generalize rule-like repetition-based patterns (such as ABB and ABA) from a sequence of elements. Increasing evidence shows that RL operates both in the auditory and the visual domain and is modulated by the perceptual expertise with the to-be-learned stimuli. Yet, whether infants’ ability to detect a high-order rule from a sequence of stimuli is affected by affective information remains a largely unexplored issue. Using a visual habituation paradigm, we investigated whether the presence of emotional expressions with a positive and a negative value (i.e., happiness and anger) modulates 7- to 8-month-old infants’ ability to learn a rule-like pattern from a sequence of faces of different identities. Results demonstrate that emotional facial expressions (either positive and negative) modulate infants’ visual RL mechanism, even though positive and negative facial expressions affect infants’ RL in a different manner: while anger disrupts infants’ ability to learn the rule-like pattern from a face sequence, in the presence of a happy face infants show a familiarity preference, thus maintaining their learning ability. These findings show that emotional expressions exert an influence on infants’ RL abilities, contributing to the investigation on how emotion and cognition interact in face processing during infancy.     

Understanding the role and function of maternal touch in children with neurodevelopmental disabilities

During the first years of life, maternal touch can serve different functions including facilitation of child’s gaze orientation to faces which is a key precursor for social attention. Although children with neurodevelopmental disability (ND) may have reduced social skills, the role of maternal touch in contributing to gaze orientation to maternal face has not been explored in previous research. In the present study, we investigate the functions of maternal touch in dyads of mothers and children with ND, compared to dyads of children with typical development (TD). Micro-analytic 2-s coding of maternal touch (i.e., affectionate, playful, facilitating, holding, no touch) and children’s gaze orientation (i.e., mother-oriented, object-oriented, non-oriented) occurred during 5-minute interactions between mothers and their 12-to-24-month-old ND and TD children. Dyads were paired for children chronological age and sex. Mothers of children with ND used more touch – especially facilitating and holding types –, compared to mothers of TD peers. Children with ND exhibited more non-orientated gaze compared to TD counterparts. Playful touch associated with increased mother-oriented gaze in children with ND, whereas affectionate and holding touch associated with greater mother-oriented gaze in TD children. Findings suggest that specific types of touch support children’s attention to social partners.     

Modeling the Influence of Language Input Statistics on Children's Speech Production

We trained a computational model (the Chunk-Based Learner; CBL) on a longitudinal corpus of child–caregiver interactions in English to test whether one proposed statistical learning mechanism—backward transitional probability—is able to predict children's speech productions with stable accuracy throughout the first few years of development. We predicted that the model less accurately reconstructs children's speech productions as they grow older because children gradually begin to generate speech using abstracted forms rather than specific “chunks” from their speech environment. To test this idea, we trained the model on both recently encountered and cumulative speech input from a longitudinal child language corpus. We then assessed whether the model could accurately reconstruct children's speech. Controlling for utterance length and the presence of duplicate chunks, we found no evidence that the CBL becomes less accurate in its ability to reconstruct children's speech with age.