Institute of Psychology, Chinese Academy of Sciences
Institutional Repository of Institute of Psychology, CASNot a member yet
28529 research outputs found
Sort by
The relationship between moral identity and prosocial behavior in adolescents: the parallel mediating roles of general belongingness and moral elevation
No full textAdolescence has long been acknowledged as a crucial period in fostering and advancing moral identity and prosocial behavior. While moral identity, in the social cognitive model, has generally been considered to promote prosocial behavior, yet a more in-depth investigation of the mediation mechanism underlying this relationship in adolescents still remains to be undertaken. Utilizing the latent variable structural equation model (SEM), this study examined both dimensions of moral identity and their relationship with prosocial behavior in adolescents (12 to 16 years), as well as the parallel mediating roles of general belongingness and moral elevation in this relationship. Participants (n = 1,000) from two public schools in China completed measures of moral identity internalization/symbolization, prosocial behavior, general belongingness, and moral elevation. Results demonstrated a positive relationship between two dimensions of moral identity and prosocial behavior. They supported that moral identity internalization and symbolization are associated with prosocial behavior through a parallel mediation effect of general belongingness and moral elevation, with moral elevation showing a stronger effect for both dimensions. In the comparison of indirect pathways, moral identity internalization demonstrated a stronger correlation with prosocial behavior than moral identity symbolization. Drawing from the social cognitive model of moral identity, this study examined the relationship between moral identity and prosocial behavior in adolescents, as well as the psychological mechanisms underlying this relationship. By examining both dimensions of moral identity and their mediating pathways to prosocial behavior, this study aimed to broaden our understanding of this relationship
Learning from financial rewards and punishments reduces the in-group bias in social approach without changing the in-group bias in impressions
No full textHumans' approach behaviour and impressions are biased towards individuals from their own group (in-group) compared with different groups (out-group). There is evidence that learning from specific interactions with in-group and out-group members can reduce these in-group biases, but it is yet unclear if learning from non-social reinforcers, such as financial rewards and punishments, can have similar effects. Here, we conducted three independent studies by using different versions of a novel approach-avoidance learning task, intergroup impression ratings and computational learning models. In the approach-avoidance learning task, participants moved a manikin representing themselves towards or away from one of two symbols, representing in-group or out-group individuals or which had no social meaning. Approach was financially rewarded with varying probabilities. Our results confirmed initial in-group biases in approach and impression ratings. Rewarding out-group approach significantly reduced the in-group bias in approach, with stronger learning from rewards compared with punishments. In contrast, the in-group bias in impressions remained unchanged. Two further studies showed that learning-related changes in approach are larger in social compared with non-social contexts and require varying reward probabilities. Together, these findings show that learning from financial rewards or punishments can improve out-group approach but not out-group impressions.</p
A comprehensive EEG dataset of laser-evoked potentials for pain research
No full textUnderstanding the neural mechanisms of pain is key to the development of novel pain diagnostic and treatment strategies. Here, we present a large-scale, comprehensive electroencephalogram (EEG) dataset of laser-evoked potentials (LEPs) from 678 healthy participants. This dataset comprises high-density EEG recordings and single-trial self-reported pain ratings in response to nociceptive laser stimuli of varying intensities (from 2.5 J to 4.5 J) delivered to either the left or right dorsum of the hand, without any physical or psychological interventions. As the largest nociceptive-evoked EEG dataset to date, it can serve as a valuable resource for mechanistic pain studies such as investigating electrophysiological underpinnings of within-individual and between-individual pain variations, for clinical studies such as providing normal values of LEPs to assess the possible lesions of pain pathways, and for methodological innovations in EEG signal processing such as optimizing preprocessing pipelines and developing new analytical tools.</p
A randomized controlled trial of a digital mindfulness intervention to improve self-handicapping in adults: serial mediating effects of mindfulness and the fear of failure
No full textSelf-handicapping poses harm to the physical and mental health, as well as personal development. The aim of the present study was to test a digital mindfulness intervention designed to reduce self-handicapping by examining mindfulness and the fear of failure as putative mediators. 73 patients with self-handicapping concerns were randomly allocated to either mindfulness (n = 31) or waiting-list group (n = 42). The intervention program had a duration of 14 days, with daily tasks requiring approximately 20-25 min to complete. Short-form five facet mindfulness questionnaire (SF-FFMQ), performance failure appraisal inventory (PFAI), and self-handicapping Scale (SHS) were measured before and after the intervention. 2 x 2 repeated-measures ANOVAs showed that compared with the waiting-list group, mindfulness significantly improved (P = 0.001), and self-handicapping and the fear of failure reduced significantly (Ps < 0.001) in the mindfulness group. Serial mediation analyses found that mindfulness and the fear of failure sequentially mediated the effects of the mindfulness invention on self-handicapping. The chain-mediated effect size accounts for 7.43% of the total effect. The digital mindfulness intervention can enhance mindfulness levels, alleviate the level of the fear of failure and self-handicapping. The chain-mediated effect of mindfulness and fear of failure may be one of the mechanisms through which digital mindfulness intervention relieves self-handicapping.</p
Marital Satisfaction Prediction Model Based on XGBoost A Study on the Trade-off Between Housework and Economic Con-tribution
No full textTraditional marital research relies mainly on questionnaires and qualitative analysis, making it difficult to effectively quantify contribution patterns and their impact on marriage. This study constructs a framework for predicting marital satisfaction based on artificial intelligence and information systems, integrating variables such as household contribution rate and family economic contribution rate, and using machine learning algorithms (XGBoost and deep neural networks) to model the data and optimise prediction accuracy. The experiment is based on marriage data collected by a distributed information system. The results show that economic contribution has a greater effect on short-term satisfaction, while household contribution has a more significant effect on long-term satisfaction. This study provides data support for electronic marriage counselling systems and promotes the development of intelligent management of marital relationships.</p
Characteristics and Mechanisms of Binocular Advantages in reading
No full text阅读作为获取信息的重要途径，在日常生活各方面都发挥着关键作用。在自然阅读情景中，大多数个体都是双眼阅读。已有研究表明正常视力个体存在阅读 双眼优势，异常视力（弱视）患者表现出阅读双眼优势受损。然而，现有研究主要聚焦于阅读速度，阅读理解层面的双眼优势尚缺乏系统探讨。同时，正常视力存在阅读双眼优势的机制和弱视阅读双眼优势受损的机制仍不明确。本论文结合心理物理学，眼动和脑电技术，从阅读速度和阅读在线理解两个维度，对比研究了正常视力和弱视的双眼优势特性及机制，涵盖 2 项研究共 8 个实验。
研究一探讨了正常视力阅读双眼优势特性及机制，包含 4 个实验。实验一通 过心理物理学方法和眼动技术探讨阅读速度双眼优势特性及行为机制。结果发现，正常视力在整体篇章阅读速度和单个汉字识别上存在双眼优势，行为机制的分析表明这种优势和正常视力双眼更好的基本视功能和眼球运动功能无关，可能是由于双眼副中央凹预加工更高效。实验二采用语义违反范式，通过眼动技术探讨阅读在线理解双眼优势特性及行为机制。结果发现，正常视力在阅读在线理解过程也存在双眼优势，体现在双眼阅读相比单眼阅读更早地探测到语义违反，这可能是由于双眼副中央凹预加工会激活语义信息。实验三和四分别采用词频效应和语 义违反范式，通过脑电技术探讨阅读速度和阅读理解双眼优势的神经机制。结果发现，双眼条件下的早期视觉成分（C1 和 P1）及晚期成分（反映词汇识别的 P2 和句子语义理解的 N400）潜伏期均早于单眼，且二者无线性相关。这表明双眼刺激会更快地激活视觉皮层及涉及词汇识别和语义整合的高级皮层，并且视觉皮层上的双眼优势效应可能不是线性地传递到高级皮层，而可能是以一种复杂的同时涉及自下而上加工和自上而下加工的多层次、分布式视觉-语言神经网络。
研究二采用和研究一相同的范式和技术探讨了弱视阅读双眼优势受损特性及机制，包含 4 个实验。实验五与六发现，弱视在阅读速度和阅读在线理解过程均不表现出双眼优势，行为机制的分析提示弱视阅读双眼优势受损可能源于弱视在基本视功能和眼球运动功能方面的双眼优势缺陷。实验七与八脑电结果显示，弱视双眼条件下的早期视觉成分（C1 和 P1）及晚期成分（P2 和 N400）潜伏期与优势眼相当，这表明弱视双眼对应的视觉通路（纹状皮层和纹外皮层）存在异常，并且影响了早期视觉加工及晚期词汇识别和句子语义理解。
综上，正常视力和弱视在阅读双眼优势特性和机制方面存在分离。在特性方面可能是量的分离，正常视力存在阅读双眼优势，而弱视则不存在显著双眼优势。在机制方面可能是质的分离，正常视力的阅读双眼优势并不依赖于基本视功能或眼球运动功能，而可能源于双眼副中央凹预加工效率和神经加工速度优势，而弱视阅读双眼优势受损来源于弱视在基本视功能和眼球运动功能的双眼优势缺陷以及弱视双眼对应的视觉神经通路存在异常。正常视力与弱视在阅读双眼优势 “特性上的量变”和“机制上的质变”揭示了两类视觉状态下双眼加工机制的系统性差异，为理解双眼视觉在阅读中的作用机制提供了更全面的视角，也为弱视 干预策略带来了新的启示。</p
Facial Video Analysis for Automated Chronic Stress Recognition: Techniques and Applications
No full text人们每天都暴露在大量信息之中，其中部分信息可能引发压力反应。适度压力是维持正常生理功能的必要因素，但长期或过度的压力会对个人、社会乃至国家造成负面影响，例如导致学习和工作效率下降，引发心理健康问题等。因此，开发有效的压力管理和干预手段至关重要。自动化压力识别技术应运而生，它通过实时监测生理信号 (如心率、皮肤电活动等) 或行为表现 (如面部表情、语音语调等) 来识别压力迹象，帮助用户及早发现并采取干预措施。此外，自动化技术还能持续追踪压力变化，帮助用户识别和应对慢性压力的累积，从而预防焦虑、抑郁等更严重问题的发生。这使得压力管理更加主动和高效，有助于人们更好地 应对现代生活中的压力挑战。然而，现有的自动化压力识别技术尚未广泛应用于日常生活中，其原因之一是实验室所诱发压力与日常生活中的压力相去甚远，想要对日常生活中的压力识别，就需要日常压力如慢性压力的数据输入。
为加速自动化压力识别技术的普及应用，本研究致力于从慢性压力的角度，提升基于面部视频的自动化压力识别模型的泛化能力，推动其更广泛的应用。本研究着重关注以下几个方面：首先，关注面部活动与慢性压力感知之间的关联，探究慢性压力状态下特定面部行为模式与不同压力水平之间的关系；其次，通过采用深度学习等多种特征提取方式构建模型，提升模型的泛化能力和鲁棒性。
因此本论文以三个部分展开。研究一旨在深入分析面部数据与心理压力之间的关系，揭示不同慢性压力水平下特定的面部行为模式；研究二旨在开发基于经典机器学习算法的自动压力评估系统，处理面部视频数据以识别压力水平，并作为基线模型用于后续比较；研究三则探索基于深度学习的面部压力自动评估方法，旨在克服传统机器学习方法在特征工程和时序建模方面的局限性，并与研究二的基线模型进行对比，验证其在提升模型性能方面的优势。研究一识别出面部信息在慢性压力感知中起关键作用的两类特征，揭示了压力水平与特定面部运动达到极值的时机之间的关系，且压力水平与某些面部部位在特定距离内的重复运动相 关。研究二在使用单变量线性回归 F 检验筛选特征的条件下，构建了类目提升回归器算法的模型，其模型皮尔逊相关系数达到了 0.539，分半信度达 0.70，表明面部视频特征能够有效用于压力评估。研究三发现 VGGISH、S3D 和 I3D 三种深度学习算法提取所得特征构建的模型，其相关系数能提升至 0.62-0.64，证明语音与视觉特征的融合能够有效提升建模精度。以上三个研究的结果表明面部信息在慢性压力评估中的潜力，且尝试多种方法有助于推动该技术的应用。
本论文的意义在于，理论上，探究了慢性压力在面部运动上的反应模式；实践应用上，利用了机器学习建模和深度学习特征提取，验证了机器学习建模能为压力与面部活动之间的映射关系，获取了基于深度学习特征提取和机器学习建模 方法表现良好的自动压力评估工具，该工具为未来基于面部数据的心理健康监测提供了可能的技术路径。</p
Multidimensional Social Network Types and Their Relation with Health Among Chinese Older Adults
No full text在老龄化程度不断加深的背景下，促进老年人健康成为重要议题。社会网络是老年人健康的影响因素之一。基于护航模型理论，社会网络涵盖结构、功能和质量三个维度(Antonucci et al., 2001, 2010)。然而研究常常使用以变量为中心的视角忽视社会网络的整合特征及老年人群体的异质性。针对上述不足，本研究使用以个体为中心的研究方法，识别老年人多维社会网络类型，关注研究中常常被忽略的功能和质量维度，并探索其随时间的变化情况。值得注意的是，对于多维社会网络类型与健康的关系，现有研究仍存在分歧，其行为、心理和生理机制仍需进一步探索。
研究一采用系统性综述的方法，根据多维度概念总结现有研究识别出的社会网络类型特点及其与健康的关联，最终纳入文献 47 篇。尽管研究普遍认为基于结构维度特征能够识别出四种典型社会网络类型(有限网络、朋友中心网络、家庭中心网络和多样网络)，但大部分研究会在此基础上发现类型的拓展和缩减，或发现非典型的丧偶组。大部分研究发现，相比于有限网络，多样网络、家庭中心网络与更好的健康状况有关，而多维度研究可能提供不支持证据。仅关注单一维度可能忽视具有组合特征的风险人群。当前多维度社会网络类别研究尚缺乏，其对健康的影响及作用机制仍有待探索。
研究二、三和四使用中国代表性纵向调查——中国健康与养老追踪调查 (China Health and Retirement Longitudinal Survey, CHARLS)，对 60 岁及以上老年人的多维社会网络类型及其与健康的关系进行探索。
研究二考察了多维社会网络类型及其转变情况。使用 2015 和2018 年两个时间点的数据，采用潜在转变分析(Latent Transition Analysis, LTA)识别社会网络类型及其在三年内的转变情况。当前研究识别出了四种社会网络类型：丧偶-高获得-高孤独、家庭中心-不满意、工作-高给予和多样-低孤独。每种社会网络类型的特点各有侧重，并非均表现为高水平或低水平。高给予他人的社会网络类型在质量维度上表现为低孤独，而高获得他人支持则表现为高孤独。大部分老年人的社会网络类型在三年的时间跨度中维持不变。
研究三检验了多维社会网络类型对健康的预测作用。使用多元回归模型、逻辑回归模型检验研究二识别出的多维社会网络类型是否能够预测 2020 年的主观健 康 、 日 常 生 活 (Activities of Daily Living, ADL) 依 赖 和 工 具 性 日 常 生 活 (Instrumental Activities of Daily Living, IADL)依赖，并且预测效果独立于 2015 年的人口统计学变量、健康变量和健康行为变量。在敏感性分析中，对缺失的数据进行了插补，在插补后的数据集中再对研究结果进行验证。结果发现，与工作-高给予相比，家庭中心-不满意组预测更低的主观健康、更高的 ADL 依赖，家庭中心-不满意组和丧偶-高获得-高孤独组预测更高的 IADL 依赖发生风险。
研究四探索了 2015 年生理因素、2018 年行为因素和 2018 年心理因素是否在多维社会网络类型和健康的关系中起中介作用。使用 Karlson/Holm/Breen (KHB)方法进行多重中介效应分析，同时控制了 2015 年的人口统计学变量和健康变量。结果发现，在社会网络类型和主观健康的关系中，抑郁起部分中介作用。在社会网络类型和 ADL 依赖的关系中，体力活动不达标和 hsCRP 异常起到部分中介作用。在社会网络类型和 IADL 依赖的关系中，抑郁起到部分中介作用。
综上，本研究使用系统性综述及以个体为中心的研究方法，揭示了多维度框架在深入理解老年人社交上的优势，发现了中国老年人多维社会网络类型特点及其在纵向上的稳定性；检验了多维社会网络类型对健康的稳定预测作用，为识别易感人群指明了方向；探索了行为、心理和生理因素在多维社会网络类型和健康之间的关系，为设计有效的干预措施奠定了基础。</p
White matter neural substrates in alcohol dependence with genetic risk and their role in pathological reward process
No full textStudies have revealed significant evidence of white matter (WM) microstructural and network connectome abnormalities in alcohol use disorder (AUD). However, the neuroimaging characteristics of alcohol dependence (AD) patients with a family history of AUD and the role of these changes in pathological craving remain unclear. Investigating the heritability of AUD is crucial for identifying genetic predispositions and informing targeted prevention strategies. The study recruited 51 patients for AD, 21 patients with a family history positive (FHP), 30 patients with a family history negative (FHN), and 25 healthy controls (HC). We compared fractional anisotropy (FA) and mean diffusivity (MD) of striatal circuits and topological properties of the reward system between the three groups. Then, covariates of alcohol use characteristics (duration and severity of AD) were controlled between FHP and FHN. We found abnormal topological properties of hippocampus in AD with FHP compared to HC. After controlling for covariates, there were still disruptions of topology organization in FHP compared to FHN, such as lower nodal betweenness, nodal degree and higher shortest path of right hippocampus. The nodal topological properties of right hippocampus were significantly correlated with self-reported craving in AD. Our findings provide robust evidence for WM neural abnormalities in AD with high genetic risk. We also found the disrupted topological properties of the right hippocampus associated with craving level.</p
Effectiveness of psychosocial interventions to improve the mental health in men who have sex with men: a systematic review and meta-analysis
No full textBackground: Men who have sex with men (MSM) are vulnerable to mental health problems. Some psychosocial interventions showed different effects on various mental health aspects, but the exact pooled effect size was uncertain. This study aimed to evaluate the effectiveness of psychosocial interventions for mental health among MSM. Methods: We included randomized controlled trials and quasi-experimental studies of psychosocial interventions aimed at improving the mental health of MSM. The outcomes were the effect sizes of overall mental health and specific aspects (depressive symptoms, anxiety symptoms, substance abuse, suicidal ideation, stress, coping, emotion, social function and identity). The fixed-effect or random-effect model was adopted to calculate the effect sizes. The study was registered with PROSPERO (CRD42024551392). Results: We included 14 studies conducted between 2010 and 2024. The effect size of intervention for overall mental health status was 0.14 (95%CI: 0.08-0.21, n=14, I2 = 28.23%). The interventions had positive effects in depressive symptoms (Hedges' g=0.25, 95%CI: 0.09-0.41), anxiety symptoms (Hedges' g=0.20, 95%CI: 0.12-0.29), substance abuse (Hedges' g=0.19, 95%CI: 0.10-0.28), stress (Hedges' g=0.18, 95%CI: 0.03-0.33), coping (Hedges' g=0.21, 95%CI: 0.06-0.36), emotion (Hedges' g=0.16, 95%CI: 0.06-0.25), and identity (Hedges' g=0.19, 95%CI: 0.07-0.30). There was no publication bias. Conclusions: Psychosocial interventions have a small-to-moderate effect on improving the mental health status of the MSM. Our study provides a comprehensive evaluation of the intervention effect, with estimations of overall mental health status and some specific aspects
Access Repository Dashboard
Do you manage Institutional Repository of Institute of Psychology, CAS? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!