作者:Zhang, Wanyi
随着大学入学考试的越来越激烈的教育竞争,高中生正面临学术和考试压力的升级,这反过来又导致了与学术情绪有关的更常见的问题。学术情绪是指学生在学习过程中经历的各种情感,与学校活动,课堂教学和学习成果直接相关1。作为学习过程中非常重要的非认知因素,学术情绪不仅会影响学生的学习成绩,而且会影响他们的心理健康发展。
佩克伦(Pekrun1。对学术情绪的研究主要集中在测试焦虑上,而其他情绪除了焦虑之外,受到较少的关注2。但是,学术情绪涵盖了各种各样的情感经历,包括愉悦,焦虑,无聊,沮丧和羞耻,所有这些学生在学习过程中都可能经历。佩克伦(Pekrun)提出的控制价值理论是研究学术情绪的至关重要的理论框架3。根据控制价值理论,控制评估和价值评估是学术情绪的关键决定因素4。控制评估是指学习者对他们对学习活动及其成果的控制程度的看法,这可以分为因果期望和因果关系5。例如,当学生认为自己努力并希望取得良好的考试成绩时,这代表了因果期望。当学生将他们的良好考试结果归因于他们的能力时,这代表了因果归因。价值评估是指学生对学习任务的价值的自我认知,可以将其分为内在的价值评估(学习者对学习任务的兴趣)和外在价值评估(学习者相信任务对他们的未来发展是有益的)。因此,学习者的学术情绪受到他们对学术活动和成果的控制和价值的看法的影响。学术情绪对学生的学习态度,学习成绩以及他们身心健康的发展产生了重大影响。
Pekrun根据价(正面或负面),唤醒水平(高或低),焦点(面向过程或以过程为导向)和时间参考(与活动相关,前瞻性或回顾性)分类学术情绪3,,,,6。这项研究采用了二维价框架,将学术情绪分为积极和负面的学术情绪。积极的学术情绪可以增强学生参与学习活动,从而增加积极的参与。尽管一些研究表明,焦虑水平中等的学生倾向于取得更好的学业成果,但由于对失败和适当的焦虑程度的恐惧会增强自我激励,并导致各种学术任务的表现提高7,,,,8,,,,9,主要的负面学术情绪的长期影响可能是有害的。从发展的角度来看,当负面情绪占上风时,他们可以减少学生对学习的热情,削弱他们的动力和兴趣,并最终对他们的学术成就产生负面影响1。
由于负面学术情绪的共同点和普遍性,这种现象通常无法吸引教育者的注意。但是,研究表明,学生在学习过程中产生的负面情绪阻碍了他们的认知发展10,负面影响他们注意力资源的分配11并限制他们的思维活动和创造力12,从而影响他们在学习方面的参与,导致学习成绩不佳和一系列心理问题。此外,负面情绪与自杀倾向有很强的相关性,表明在青春期中经历这种情绪可能会提高患上心身疾病的风险,并在成年后期的学术调整中面临挑战13,,,,14。因此,关注和研究高中生的负面情绪具有重要的实际意义。
Pekrun的社会认知模型强调了个人认知评估和社会环境对学术情绪的重大影响。基于Pekrun的理论框架,本研究将个体的认知变量视为负面学术情绪的预测指标,同时将环境影响视为重要的上下文因素。社会认知模型通过探索学生如何处理学习环境以了解他们的情感经历,为分析负面的学术情绪提供了明确的结构途径。根据发展情境主义,个人情感的发展是个人与环境之间不断互动的结果15。与强调主观认知评估的Pekrun的模型相反,发展情境提供了更广泛的发展系统观点。它强调,高中生的负面学术情绪是由个人特征和上下文因素的共同影响而产生的。在强调环境的作用的同时,该理论也强调了稳定的个人特征的重要性。它为本研究提供了一种新的观点,建议除了认知因素外,还应考虑个人在学习过程中表现出的稳定趋势。
这两个理论框架是互补的,因为既突出了内部和外部因素对情绪发展的互动影响。但是,他们对内部因素的重视有所不同。社会认知理论的重点是个人认知过程在与环境互动期间塑造情感经历中的作用,强调诸如自我效能和归因风格之类的结构。相比之下,发展情境主义结合了相对稳定的个体特征,例如心理弹性,随着时间的流逝,它们与情境因素动态相互作用。基于这两种理论,这项研究将导致高中生的负面情绪的因素分为两类:个人因素和背景因素。选择影响高中生学术情绪的个体因素包括心理韧性,归因风格和自我效能感。主要背景因素是学校环境,重点是教师的纪律风格。这两种理论是互补的,强调了个人发展与环境之间的相互作用。
学校是学生学习和生活的主要环境,它也是影响个人的重要因素之间的关键微观系统。应该在学校背景下对高中生负面的学术情绪进行研究。老师是学生学习和生活的主要指南。先前的研究表明,不同的学科风格会影响学生的心理学和行为16,,,,17。权威和公平的纪律风格倾向于引起学生的积极情绪,导致更积极的行为,而冷漠的纪律风格可能会产生负面情绪,并触发更多的负面行为18。
先前的研究表明,个人的适应能力,尤其是通过心理韧性表现出来,在面对挫折和挑战时在减轻负面的学术情绪中起着至关重要的作用19。心理韧性是指面对逆境,创伤,悲剧,威胁或其他重大生活压力的个人的积极适应20。它反映了一个人从压力和挫折中恢复的能力。遇到困难或失败时,心理韧性低的学生通常会表现出不良的情绪调节21。他们倾向于缺乏有意义的人际关系,努力寻求帮助或表达自己的情绪,并难以有效应对生活或学习中遇到的挑战。结果,它们更容易受到负面学术情绪的破坏性影响22。
除了心理上的韧性外,个人的归因风格是影响认知评估的另一个重要因素。归因是指人们对事件原因的解释和评估,归因样式是指个人在个人多次归因后形成的属性的稳定趋势23。归因理论指出,事件结果的归因会影响情绪,而归因的方式显着影响个人的情绪24。对于学生来说,归因可以有效地预测学术情绪。一种归因于内部成功和外部失败(将成功归因于自己和外部因素失败)的归因风格可以产生积极的学术情绪25,虽然在外部和内部归因于失败(将成功归因于外部因素和失败)可能导致负面的学术情绪26。研究证明,学生的学术情绪通过归因培训得到了改善,进一步证实了归因在学术情绪中的预测作用27。
此外,研究证实了学术自我效能和学术情绪之间的相关性。学术自我效能是指个人完成学术任务能力的判断和评估28。调查发现,学术自我效能水平较低的学生更有可能在学习中经历负面的学术情绪29,,,,30,,,,31,,,,32。一项讨论数学学习自我效能和数学学术情绪之间关系的研究发现,数学学习自我效能可以显着预测数学学术情绪33,在英语学习的背景下获得了相似的结果34。随着计算机网络的发展,许多学者研究了在线学习中的学术自我效能和学术情绪之间的关系,发现在线学习自我效能仍然在预测学术情绪中起着重要作用35,,,,36。
先前的研究已经确定了与学生的负面学术情绪有关的许多因素,例如父母的参与37,教师支持38和学生的心理健康39。但是,传统的相关研究通常无法提供早期预测模型40。此外,缺乏包括不同心理因素的综合框架限制了研究发现的实际应用。机器学习是一种基于用于检测数据中隐藏模式的算法和统计模型的方法,可以通过交叉验证优化分析结果,并且受异常值的影响较小41。它已成为教育数据挖掘中最受欢迎的工具之一,在学生筛查和干预等领域显示出巨大的潜力42。它在预测学生的自杀念头,自我伤害行为方面取得了积极的成果43,但迄今为止,还没有使用机器学习来预测学术情绪。因此,本研究的目的是:(1)使用机器学习算法建立一个有效的高中生学术情绪的预测模型;(2)识别和分析影响这些情绪的关键因素;(3)促进具有高水平的负面学术情绪的学生的早期识别和干预,最终为他们的心理健康和学术上的成功做出了贡献。
这项研究涉及从河北省的几所普通高中选择一年级和二年级的高中课程。问卷是按照纸张格式分发的,并在常规课程期间由学生匿名完成。总共分发并收集了1708个问卷。所有参与者都符合中国年龄的入学要求,该年龄在15至17岁之间。其中有749名男学生(43.9%)和959名女学生(56.1%);就年级而言,有745名高中生(43.6%)和963年的高中生(56.4%);关于家庭结构,只有984名儿童(57.6%)和724名非仅儿童(42.4%)。所有接受调查的学生均签署了知情同意书。在删除缺少值的受试者后,总共保留了1696个有效样本,实现了99.3%的利用率。在官方调查之前,这项研究获得了科学研究伦理委员会的许可。所有参与者及其监护人获得了知情同意,以确保他们对调查目的的理解。调查说明清楚地说明了任何模棱两可的项目,并强调了调查的自愿性和机密性。
教师纪律风格通过教师纪律风格量表进行评估44。该问卷分为两个分量表:教师回应和教师要求,共有17个项目。教师的反应量表包含9个项目(例如,当我表现不佳时,老师会安慰我),并且老师要求的量表包含8个项目(例如,当我在考试中表现不佳时,老师在课堂上责骂我的责任),从未将教师从未划分至5个倾向。这两个子量表的Cronbach alpha系数分别为0.92和0.88。
为了衡量弹性,采用了心理弹性量表45,其中由27个以5分制的项目组成(从未有1个几乎从不到5点)。它包括两个维度:个人和支持力。个人进一步分为三个因素:目标计划(例如,我一生中有明确的目标),影响控制(5个项目,例如,我很难控制不愉快的情绪)和积极的思维方式,和积极的思维方式(4个项目,例如,我相信逆境可以激励我的两项,同时支持我的两项。寻求帮助(6个项目,例如,我遇到困难时与他人交谈)。更高的分数表明该项目的心理韧性水平更高,需要12个项目需要反向评分。在这项研究中,量表的Cronbach的α系数为0.90。
归因方式是使用多维多属性因果关系量表(MMCS)MMC的学术成就量表进行的46。MMC使用5点评级量表(从1个完全不同意到5完全同意)。It includes four dimensions: attribution to ability (6 items, e.g., âThe most important factor in achieving good grades is my learning abilityâ), attribution to effort (6 items, e.g., âPoor grades indicate that I didnât work hard enoughâ), attribution to background (6 items, e.g., âSometimes I get good grades just because the subject is easy to learnâ), and attribution to luck (6 items, e.g.,有时候,我因为有些运气而成功。该量表的Cronbach的αα系数为0.75。
通过学术自我效能调查表评估了学术自我效能调查表47。该问卷由22个额定级别的项目组成(从1个完全不同意到完全同意)。它包括两个维度:学习能力的自我效能感(例如,11个项目,我相信我有能力在学习中取得良好的成果)和学习行为的自我效能感(例如,11个项目,例如,我总是突出教科书或笔记本中的关键部分以帮助学习学习)。该量表的Cronbach的Alpha系数为0.83。
通过缩短的学术情绪问卷(S-EES)的缩短版本检查了负面的学术情绪48。问卷包含七个项目:惊喜,好奇心,兴奋,焦虑,混乱,挫败感和无聊,并以5分制(从1个完全到5点至5点)进行评分。负面的学术情绪包括焦虑,混乱,沮丧和无聊。在数学学习情绪的研究中,发现“无聊”对两个因素的负载相同,这些因素不符合测量标准,因此不包括在负面情绪中49。
鉴于张的研究仅关注数学,因此其发现可能不适用于本研究。因此,这项研究对问卷进行了验证性因素分析。结果表明,两因素模型拟合得很好(•3.34,srmrâ= 0.04,gfiâ= 0.97,cfi = 0.99,ifi = 0.99,tli = - 0.98)。负面学术情绪中无聊的因素负载(0.78)高于积极的学术情绪(0.41),因此,在这项研究中,无聊归因于负面的学术情绪。问卷的内部一致性为0.93。
(1)数据转换。为了避免由于不同类型的变量之间的数值差异而导致的预测错误,进行了数据归一化50,标准数据的平均值为0,标准偏差为1。(2)数据集除法。根据Zâ> 1和Zâ€1的标准,主体被分为高级和中低级小组,内容涉及负面的学术情绪51。采用了10倍的交叉验证,这意味着原始数据集分为10个大约相等的子集。在每次迭代中,将9个子集用作培训数据,其余的数据作为验证数据。这个过程重复了10次。在每次迭代中,对模型进行了不同的培训数据培训,然后对相应的验证数据进行了评估。这种验证方法准确反映了模型的概括能力,并提供了更好的操作效率和稳定性52。(3)数据处理不平衡。由于样本数据集偏斜,导致机器学习预测偏向多数级别,因此使用合成的少数群体过采样技术(SMOTE)进行过度采样53。该技术通过生成少数族裔类别的合成样本来帮助平衡数据集。
使用SPSS 22.0分析数据,以进行描述性统计,统计推断和常见方法偏差的测试。Harman单因素测试方法用于常见方法偏置测试。通过该测试的标准至少具有两个因素,其特征值大于1,最大因素解释了差异的40%。
使用贝叶斯线性混合效应模型检查了组差异。这种建模方法由于其在贝叶斯统计推断中的优势而引起了人们的关注,即,将先前信息与模型可能性相结合的能力估算后验分布及其处理复杂层次结构的能力54,,,,55。使用BRMS软件包在R中进行了分析。该模型为响应变量指定了一个学生()分布,以提高针对重尾数据的鲁棒性。使用四个马尔可夫链进行参数估计,每个链带有4,000个迭代,其中前1,000个迭代用于热身。这导致了总共12,000个后部样本。提取汇总系数(估计值),95%贝叶斯可靠间隔(CIS),自由度(½)和马尔可夫链蒙特卡洛(MCMC)收敛诊断(RHAT)的摘要统计数据。为了进一步检查群体差异,使用BRMS软件包提供的假设()函数计算了贝叶斯因子56。
对于模型构建,使用了基于Scikit-Learn版本1.1.0的Python 3.8软件。我们采用了各种算法,包括逻辑回归(LR),Naive Bayes分类器(NBC),支持向量机(SVM),决策树(DT),Random Forest(RF),梯度增强决策树(GBDT)(GBDT)和适应性提升(Adaboost)。这些模型在处理分类任务中的有效性已得到广泛验证。我们比较了这七个模型的性能,并根据结果选择了最佳模型。在比较了跨模型的结果之后,选择了最佳总体绩效的结果以进行进一步的分析和解释。该模型是预测负面学术情绪并探索其潜在影响因素的基础。
使用四个通常采用的指标评估模型的性能:准确性,精度,召回率以及接收器操作特征曲线(AUC)下的面积。精度是指正确预测的样本总数中正确预测的结果的百分比,从而提供了模型的预测性正确性的总体度量。精度是指在模型中预测的所有属于高消极学术情绪组的学生中,在高负学术情绪组中的比例。高精度意味着大多数被预测处于高风险的学生确实面临着严重的负面学术情绪问题。记录表明,在高负学术情绪组中的学生比例是该模型成功预测的。高回忆意味着该模型可以全面地确定所有面临负面学术情绪的高风险的学生。AUC是机器学习中广泛使用的评估指标,尤其是在二进制分类任务中。它根据预测的概率或分数来评估模型的性能在区分正面和负样本时。更高的AUC值表示模型更好的判别能力,其AUC值较大,表示更好的分类效果。为了全面评估模型的精确度和回忆,本研究报告了F1分数。F1分数提供了一个单一的度量,该指标可以平衡两者之间的权衡,尤其是在误报和假否定的情况下,要考虑的情况至关重要。
由于这项研究采用了10倍的交叉验证方法来训练模型,因此将获得十个精度,召回和AUC值,最终结果是这十个值的平均值。这种方法可确保对数据的不同子集对模型的性能进行更可靠,更广泛的评估。
使用Harman单因素分析方法来测试数据中常见方法偏差。结果表明,特征值大于1的30个因素,最大因素占差异的16.31%(小于40%)。因此,这项研究不会遭受严重的常见方法偏见。
如表所示 1,负面学术情绪的平均得分为12.99,比理论中位数略高,这表明高中生的负面学术情绪处于高于平均水平的水平。就教师学科而言,教师反应的平均得分远高于理论中间,而教师需求的平均得分明显低于理论中位数。这意味着教师在管理学生时倾向于倾向于沟通而不是控制。高中生的心理韧性的平均得分为93.56,表明高水平的心理韧性。学术自我效能感的平均得分为73.22,表明高中生的自我效能水平高于平均水平。
中低组总共包括1460个个体,而高层组成236个人。贝叶斯混合效应建模的结果显示在表中 2。结果表明,对于心理弹性,不可控制的因素归因(能力,背景和运气),学术自我效能感和教师纪律风格的95%可靠间隔(CI)并未包括零及其相应的贝叶斯因素(BFâ> 3),提供了两组之间的大量差异。相比之下,努力归因的95%CI包括零,贝叶斯因子小于1,表明两组之间没有有意义的差异,中等证据支持这种无效效应。所有RHAT值均低于1.01,表明模型的良好收敛性。此外,许多变量的自由度(½)低于30,支持了指定的适当性学生响应变量的分布。
使用13个变量,包括心理弹性,归因风格,学术自我效能和教师学科样式作为自变量,以及将高中生负面学术情绪作为因变量的分类,用于高中生的负面学术情绪水平的预测模型,使用逻辑回归(LR),Naive Bayes Classifier(NB)(NB)(NB)(nbc),Support vector(nbc),Support vector(S)(RF),梯度提升决策树(GBTD)和Adaboost。这七种方法能够在一定程度上预测高中生的负面学术情绪水平。在对精度,召回,F1得分和AUC进行全面比较后,随机森林算法在所有指标中表现出最佳性能。见表 3有关详细信息。表3学术情绪水平的不同模型的预测性能(
根据图 1,情绪控制在预测负面的学术情绪方面做出了最大的贡献。排名最高至最低的五个因素是影响控制,研究能力自我效能,归因于运气,归因于背景以及学习行为自我效能。
The aim of this study was to establish an efficient predictive model using various machine learning algorithms, thereby assisting educators in timely identifying and intervening with high school students who have high levels of negative academic emotions.The results show that all the machine learning models selected for this study demonstrated a certain degree of effectiveness in predicting negative academic emotions among high school students, with all models achieving an AUC value greater than 0.7.This outcome further illustrates the potential application of machine learning technology in the field of educational psychology.
Among all models, the random forest model showed significant superiority in this study, with an AUC value reaching up to 0.96 and precision exceeding 80%, far surpassing other machine learning models.Compared to other models, the random forest model has a unique construction method and algorithmic principle.Random forest is an ensemble learning method that improves the overall predictionâs accuracy and stability by building multiple decision trees and aggregating their predictions.When dealing with complex, high-dimensional educational data, random forest effectively reduces data noise and avoids overfitting, thereby more accurately capturing the factors that affect studentsâ negative academic emotions.Additionally, the random forest model enhances prediction accuracy by evaluating the voting outcomes of different decision trees.This method gives the model strong robustness when facing large-scale and diverse datasets.
The development of negative academic emotions is a complex psychological process, not caused by a single factor, but rather the result of the interplay of multiple factors.This study, utilizing machine learning methods, especially the random forest algorithm, effectively analyzed the complex relationships among these factors, providing a more comprehensive perspective for understanding the intricate mechanisms behind negative academic emotions.By taking into account variables across multiple dimensions, including individual psychological resilience, attribution style, and academic self-efficacy, this study has revealed the core factors influencing negative academic emotions in high school students.
In RF, emotional control, self-efficacy in learning ability, attribution to luck, attribution to background, and self-efficacy in learning behavior were identified as the top five key factors for predicting the level of negative academic emotions.These variables are distributed among the three factors previously selected for the study: psychological resilience, attribution style, and academic self-efficacy.They reflect how students perceive and respond to learning challenges and stress, as well as how they evaluate their own learning capabilities and achievements.
In this study, all dimensions of psychological resilience played a certain predictive role in forecasting negative academic emotions among high school students, with emotional control identified as the most significant contributing factor among all predictive variables.This result is consistent with previous research.Students with higher levels of psychological resilience are more likely to experience positive academic emotions, such as happiness, whereas those with lower levels of psychological resilience are more prone to negative emotions, such as anxiety52。This confirms that psychological resilience is an important factor in academic emotions.Building on previous research, this study further verifies the critical role of emotional control in negative academic emotions.Emotional control refers to studentsâ ability to adjust and control pessimistic emotions in difficult situations45。Whether students are adept at controlling their emotions predicts negative academic emotions.Academic emotions, as a sub-concept of emotions, represent a specific domain of emotional response, reflecting studentsâ emotional experiences related to the learning process and outcomes.If high school students cannot effectively control and regulate their emotions, they are more likely to experience negative emotions such as frustration, boredom, and despair when encountering setbacks and difficulties in learning.These negative academic emotions can not only affect studentsâ learning motivation and academic achievement but may also have adverse effects on their long-term psychological health.
This study examined the impact of attribution style on negative academic emotions among high school students, highlighting the significant role of background and luck attributions in predicting negative academic emotions.Weinerâs theory of motivation and attribution suggests that individualsâ attributions for success or failure directly affect their emotional responses.Studies have shown that there is a significant positive correlation between extrinsic learning motivation and negative academic emotions.Individuals with strong extrinsic motivation experience greater learning pressure, leading to psychological fragility and high levels of stress, which in turn generate negative emotions57。Building on this, the results of the current study further emphasize the critical role of uncontrollable attributions in predicting negative academic emotions.Attributions to ability, background, and luck all have significant predictive effects on negative academic emotions, whereas the predictive effect of effort attribution is not as apparent.This reflects the unique role of effort attributionâdirectly affecting academic achievement, while other attribution tendencies impact academic achievement through academic emotions as a mediator58。
This study found academic self-efficacy to be a critical predictive factor, in agreement with earlier research findings that highlight its important role in shaping studentsâ academic self-concept59。Academic self-efficacy encompasses two dimensions: self-efficacy in learning ability and self-efficacy in learning behavior, reflecting studentsâ confidence in their learning capabilities and their sense of control over learning outcomes, respectively.
According to self-worth theory, negative self-assessments of oneâs abilities can lead to a compromised sense of self-worth, subsequently triggering feelings of shame60。In an academic context, if students lack confidence in their learning abilities, they might feel inadequate to meet learning challenges.This perception can not only impact their academic performance but also lead to negative emotional experiences.Furthermore, control-value theory suggests that when individuals feel they have lost control over learning outcomes, their expectations for the future diminish, potentially giving rise to feelings of boredom.Self-efficacy in learning behavior reflects studentsâ beliefs in their control over the learning process and outcomes59,,,,61, where a high level of self-efficacy in learning behavior helps students maintain a positive attitude towards learning activities and high engagement.Conversely, when students feel they cannot effectively control the outcomes of their learning, they may experience feelings of helplessness and frustration, leading to disinterest and negative emotions towards learning39,,,,62。
This research contributes to understanding the influence of individual and environmental factors on negative academic emotions in high school students from the perspective of complex models and reveals the feasibility and potential significance of predicting academic emotions through machine learning algorithms.In educational practice, teachers and schools can leverage these findings to implement more personalized interventions aimed at helping students cope with academic stress.By guiding students to develop emotional regulation skills and adopt more adaptive attribution styles, the adverse effects of negative academic emotions on academic performance and mental health can be effectively mitigated.Moreover, the application of machine learning techniques offers educators a valuable tool for predicting studentsâ emotional states, enabling early warning and timely intervention.This has the potential to enhance studentsâ academic performance and psychological adjustment.
This study has several limitations.Firstly, this study utilized self-report questionnaires as research tools, which might be subject to social desirability bias as participants could respond in a manner they believe is expected by society.Additionally, the use of cross-sectional analysis makes it challenging to track the temporal changes in how various factors influence negative academic emotions.Furthermore, the selection of factors in this study is not comprehensive, failing to cover all possible elements that could affect studentsâ academic emotions.In addition, the sample was limited to first- and second-year high school students in Hebei Province, China.Given that educational policies in China are uniformly guided by the Ministry of Education, the findings may be generalizable to other regions within the country.However, their applicability to different educational stages and cultural contexts warrants further investigation.Based on the limitations of this study, future research can be expanded in several directions.First, future studies may consider incorporating a wider range of factors that potentially influence studentsâ academic emotions and utilize longitudinal data to explore the long-term trends of emotional changes caused by these factors.Second, further research is needed to validate the adaptability of machine learning models across different cultural contexts and educational systems.Third, by integrating specific educational intervention strategies, future studies could examine whether emotion management approaches based on predictive analytics can effectively reduce negative academic emotions in practical applications.Research may also explore the integration of machine learning with traditional educational psychology methods to identify more efficient intervention techniques.This approach would enhance the capacity to design targeted interventions and offer greater potential for mitigating studentsâ negative academic emotions.
In conclusion, this study is the first to develop a predictive model of negative academic emotions among high school students using machine learning algorithms.Among the various algorithms tested, the random forest model exhibited superior predictive accuracy.The variable importance analysis indicated that emotional control and attributional style significantly contributed to predicting studentsâ negative academic emotions.These findings provide actionable insights for educators and policymakers.For example, educators could effectively reduce studentsâ negative academic emotions by cultivating positive attributional patterns during learning activities.Additionally, schools may implement targeted intervention programs aimed at enhancing studentsâ psychological resilience and improving their ability to cope with setbacks and difficulties, thus promoting overall psychological well-being and academic achievement.
根据合理的请求,在当前的研究中使用和/或分析的数据集使用。
Pekrun, R., Goetz, T., Titz, W. & Perry, R. P. Academic emotions in studentsâ self-regulated learning and achievement: A program of qualitative and quantitative research.教育。Psychol。 37, 91â105 (2002).
文章一个 Google Scholar一个
Frenzel, A. C., Pekrun, R. & Goetz, T. Perceived learning environment and studentsâ emotional experiences: A multilevel analysis of mathematics classrooms.学习。操作说明。17, 478â493 (2007).文章
一个 Google Scholar一个 Pekrun, R. The control-value theory of achievement emotions: assumptions, corollaries, and implications for educational research and practice.Educational Psychol.
修订版18 , 315â341 (2006).文章
一个 Google Scholar一个 Pekrun, R., Frenzel, A. C., Goetz, T. & Perry, R. P. The control-value theory of achievement emotions: an integrative approach to emotions in education inEmotion Education
(eds. Schutz, P. A. & Pekrun, R.), 13â36 (Elsevier, 2007).Yao, W. J., Jiang, Q., Li, Y. & Zhao, W. Research on influencing factors and science of science of wakeup mechanism of academic emotion under crowdsourcing knowledge ConstructionââResearch on the structural change of classroom teaching for deep learning (V).Modern Distance Education.
5, 33â42 (2020).Pekrun, R. et al.A three-dimensional taxonomy of achievement emotions.
J. Personal.Soc。Psychol。 124, 145 (2023).
文章一个 Google Scholar一个
Al-Qaisy, L. M. The relation of depression and anxiety in academic achievement among group of university students.int。J. Psychol。Counselling。3, 96â100 (2011).Google Scholar
一个 Andrews, B. & Wilding, J. M. The relation of depression and anxiety to life-stress and achievement in students.
br。J. Psychol。 95, 509â521 (2004).
文章一个 PubMed一个 Google Scholar一个
Eysenck, M. W., Derakshan, N., Santos, R. & Calvo, M. G. Anxiety and cognitive performance: attentional control theory.情感 7, 336 (2007).
文章一个 PubMed一个 Google Scholar一个
Efklides, A. & Volet, S. Emotional experiences during learning: multiple, situated and dynamic.学习。操作说明。15, 377â380 (2005).文章
一个 Google Scholar一个 Meinhardt, J. & Pekrun, R. Attentional resource allocation to emotional events: an ERP study.Cogn。
Emot.17 , 477â500 (2003).文章
一个 PubMed一个 Google Scholar一个 Lu, J. M. Classification of affective goals of classroom teaching.J. Psychol。
科学。30 , 1291â1295 (2006).Google Scholar
一个 Copeland, W. E., Shanahan, L., Costello, E. J. & Angold, A. Childhood and adolescent psychiatric disorders as predictors of young adult disorders.
拱。Gen. Psychiatry。66, 764â772 (2009).文章
一个 PubMed一个 PubMed Central一个 Google Scholar一个 Peng, N., Luo, N., Zhu, C. Y., Zhou, W., Gao, G. D. & Y. F. & Risk behaviours of adolescents in Shanghai.Shanghai J. Prev.
医学4 , 163â167 (2003).Google Scholar
一个 Zhang, W. X. & Chen, G. H. Developmental contextualism:an instance of development system theories.
ADV。Psychol。科学。 17, 736â744 (2009).
Alamgir, M., Ali, M. & Rehman, A. Influence of teacher and student relationship on the academic motivation in the presence of academic emotions.J. Dev。社会科学。 5, 64â74 (2024).
Handayani, A. R., Mairanda, J., Giartriweni, N. K., Nirvananda, W. K. & Wijaya, E. The role of humor style and emotional intelligence on academic stress in adolescents.Edunity Kajian Ilmu Sosial Dan.Pendidikan。3, 103â112 (2024).文章
一个 Google Scholar一个 Chen, J. & Zhang, J. J. Perception Emotion experience and behavior tendency toward the message of Teacherâs disciplines.J. Psychol。
科学。35, 344â346 (2004).Zhao, F. Q. & Yu, G. L. Everyday academic resilience: active adaption to everyday academic pressures.
ADV。Psychol。科学。 26, 1054â1062 (2018).
文章一个 Google Scholar一个
Hu, Y. Q. & Gan, Y. Q. Development and psychometric validity of the resilience scale for Chinese adolescents.Acta Physiol.西尼卡。40, 902â912 (2008).Google Scholar
一个 Mei, Z. L. & Liao, C. G. Psychological resilience and emotional self-regulation of higher vocational college students.
环境。Social Psychol. 9, 6213 (2024).
文章一个 Google Scholar一个
Jiang, Z. C. & Xu, Z. J. The relationship among middle school studentsâ academic emotions,resilience and school-work achievement.China J. Health Psychol. 25, 290â293 (2017).
Zhang, T. & Zhou, M. The affiliation attribution style of medical college students.China J. Health Psychol. 23, 1510â1512 (2015).
Hailikari, T., Nieminen, J. & Asikainen, H. The ability of psychological flexibility to predict study success and its relations to cognitive attributional strategies and academic emotions.Educational Psychol. 42, 626â643 (2022).
文章一个 Google Scholar一个
Stiensmeier-Pelster, J. & Heckhausen, H. Causal attribution of behavior and achievement inMotivation and Action(eds. Heckhausen, J. & Heckhausen, H.) 623â678 (Springer, 2018).
Erkut, S. Exploring sex-differences in expectancy, attribution, and academic-achievement.性别。角色。9, 217â231 (1983).文章
一个 Google Scholar一个 Ma, H. X. & Zhang, H. An experimental research on improving studentsâ academic emotions by attribution training.理论实践。
教育。33 , 37â39 (2013).Google Scholar
一个 Diao, C. T., Zhou, W. Q. & Huang, Z. The relationship between primary school studentsâ growth mindset, academic performance and life satisfaction: the mediating role of academic Self-Efficacy.
螺柱。Psychol。行为。 18, 524â529 (2020).
Putwain, D., Sander, P. & Larkin, D. Academic self-efficacy in study-related skills and behaviours: relations with learning-related emotions and academic success.br。J. Educ.Psychol。 83, 633â650 (2013).
文章一个 PubMed一个 Google Scholar一个
Villavicencio, F. T. & Bernardo, A. B. I. Negative emotions moderate the relationship between Self-Efficacy and achievement of Filipino students.Psychol。螺柱。 58, 225â232 (2013).
Lin, J., Liu, Y. L. & Peng, W. B. The relationship between college studentsâ academic emotion and learning engagement: the mediating role of academic Self-Efficacy.下巴。J. Special Educ. 4, 89â96 (2020).
Tang, L. & Zhang, S. Academic emotion and its relationship with academic self-efficacy among secondary special school students.下巴。J. Behav.医学脑科学。 18, 456â458 (2009).
Jiang, S. Y., Liu, R. D., Zhen, R., Wei, H. & Jin, F. K. Relations between fixed mindset and engagement in math among high school students: roles of academic Self-efficacy and negative academic emotions.Psychol。开发教育。 35, 48â56 (2019).
Zhang, D.A Research of High School Studentsâ English Academic Emotions and their Relations with English Academic self-efficacy(Henan University, 2010).
Deng, W. B. et al.Effects of regulatory focus on online learning engagement of high school students: the mediating role of self-efficacy and academic emotions.J. Comput。协助。学习。 38, 707â718 (2022).
文章一个 Google Scholar一个
Wang, Y. Q. et al.Interaction and learning engagement in online learning: the mediating roles of online learning self-efficacy and academic emotions.学习。Individual Differences。94, 10 (2022).文章
一个 Google Scholar一个 Xu, X. K., Deng, C. P. & Liu, M. ParentsâAcademic involvement and negative emotions in high school students:the mediating role of Parent-Child relationship and the moderating role of parental psychological control.J. Psychol。
科学。43 , 1341â1347 (2020).Google Scholar
一个 Gao, Y. et al.
Teacher support and emotional experience of middle school students:a moderated mediating model.China J. Health Psychol. 29, 629â634 (2021).
Wang, D. Y. & Zhou, L. Between psychological well-being and negative academic emotions in left-behind middle school student in Anhui Province in 2016: the moderating role of cognitive reappraisal.J. Hygiene Res. 46, 935 (2017).
Ding, X. F., Nie, J. & Zhang, B. Using demographic information, psychological assessment data and machine learning to predict studentsâ academic performance.J. Psychol。科学。 44, 330â339 (2021).
Gowin, J. L. et al.Using neuroimaging to predict relapse in stimulant dependence: A comparison of linear and machine learning models.NeuroImage-Clin 21, 7 (2019).
文章一个 Google Scholar一个
Yarkoni, T. & Westfall, J. Choosing prediction over explanation in psychology: lessons from machine learning.Perspect.Psychol。科学。 12, 1100â1122 (2017).
文章一个 PubMed一个 PubMed Central一个 Google Scholar一个
Ma, M., Liu, H. & Liu, R. X. Application of machine learning in the prediction of college studentsâsuicidal ideation.下巴。J. School Health。43, 763â767 (2022).Google Scholar
一个 Sun, M. Y. & Wang, Z. H. The development and confirmatory factor analysis of teacher discipline scale for junior high school students.
Psychol。测试。 55, 611â633 (2008).
Hu, Y. Q. & Gan, Y. Q. Development and psychometric validity of the resilience scale for Chinese adolescents: development and psychometric validity of the resilience scale for Chinese adolescents.Acta Physiol.西尼卡。40, 902â912 (2008).Google Scholar
一个 Li, C. X.
Research on Resilience, Attributional Style and Academic Selfâ-efficacy for Students Attending College Entrance Exam(Jilin University, 2014).Liang, Y. S.
Study on Achievement Goal, Attribution Styles and Academic Self-efficacy of Collage Students(Central China Normal University, 2000).Pekrun, R., Vogl, E., Muis, K. R. & Sinatra, G. M. Measuring emotions during epistemic activities: the Epistemically-Related emotion scales.
Cognition Emot.31 , 1268â1276 (2017).文章
一个 Google Scholar一个 Zhang, Y. Y.The Influence Mechanism of Mathematics Academic Emotion on Mathematics Achievement of Senior High School Students and its Intervention Study
(Hebei Normal University, 2022).Tezcan, J. & Cheng, Q. Support vector regression for estimating earthquake response spectra.公牛。
Earthq.工程。 10, 1205â1219 (2012).
文章一个 Google Scholar一个
Lin, X., Huang, Y. L., Yangh, J. & Tang, P. Analysis of employment anxiety factors of medical students based on random forest algorithm.J. Chengdu Med.科尔。 17, 764â768 (2022).
Liang, Z. C., Li, Z. Y. & Lai, Q. Application of 10-fold cross-validation in the evaluation of generalization ability of prediction models and the realization in R.下巴。J. Hosp.Stat. 27, 289â292 (2020).
Chawla, N. V., Bowyer, K. W., Hall, L. O. & Kegelmeyer, W. P. SMOTE: synthetic minority over-sampling technique.J. Artif.Intell。res。 16, 321â357 (2002).
文章一个 数学一个 Google Scholar一个
Bürkner, P. Brms: an R package for bayesian multilevel models using Stan.J. Stat。软件。 080, 1â28 (2017).
文章一个 Google Scholar一个
Sorensen, T. & Vasishth, S. Bayesian linear mixed models using Stan: A tutorial for psychologists, linguists, and cognitive scientists.量子。Methods Psychol. 12, 175â200 (2015).
文章一个 Google Scholar一个
Pan, W. K., Wen, X. J. & Jin, H. Y. Bayesian Mixed-effects models: A primer with Brms.Psychology:Techniques Appl. 11, 577â598 (2023).
Li, L. I. Relationship between learning motivations and academic emotions among medical freshmen.mod。上一条。医学 40, 497â499 (2013).
广告一个 Google Scholar一个
Chen, J. J. & li, S. F. The Paths of Junior School Studentsâ Achievement Attribution and Academic Emotions Forecasting Their Academic Achievement.Chinese J. Clin.Psychol。 20, 392â394 (2012).
Wang, D. Y., Lu, X. & Yin, X. The association of negative academic emotions on perceived academic Self-efficacy of migrant children: the moderating role of emotion regulation strategies.Psychol。开发教育。 33, 56â64 (2017).
Covington, M. V. Goal theory, motivation, and school achievement: an integrative review.安。Rev. Psychol. 51, 171â200 (2000).
文章一个 Google Scholar一个
Jiang, G. R. et al.The status quo and characteristics of Chinese mental health literacy.Acta Physiol.西尼卡。53, 182â198 (2021).Google Scholar
一个 Pekrun, R., Elliot, A. J. & Maier, M. A. Achievement goals and discrete achievement emotions: A theoretical model and prospective test.
J. Educ.Psychol。 98, 583â597.https://doi.org/10.1037/0022-0663.98.3.583(2006)。
文章一个 Google Scholar一个
We are grateful to all the participants in this study.
This work was supported by the Post-Funded Projects of The National Social Science Fund(No.22FJKB019).
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethics Committee of the School of Psychology of Hebei Normal University (ID of approval: LLSC2024060).
作者没有宣称没有竞争利益。
The informed consents were obtained from the participant and their parents or legal guardians.
作者没有宣布的利益冲突。
关于已发表的地图和机构隶属关系中的管辖权主张,Springer自然仍然是中立的。
Below is the link to the electronic supplementary material.
开放访问This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material.您没有根据本许可证的许可来共享本文或部分内容的改编材料。The images or other third party material in this article are included in the articleâs Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the articleâs Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.要查看此许可证的副本,请访问http://creativecommons.org/licenses/by-nc-nd/4.0/。重印和权限
等。Constructing a predictive model of negative academic emotions in high school students based on machine learning methods.Sci代表15 , 19183 (2025).https://doi.org/10.1038/s41598-025-04146-6
已收到:
公认:
出版:
doi:https://doi.org/10.1038/s41598-025-04146-6