数字化认知评估工具的开发及其在儿童群体中的应用

Development of digital cognitive assessment tools and its application in children

中文关键词:  数字化认知评估工具  认知功能  儿童

英文关键词:Digital cognitive assessment tool  Cognitive function  Children

中文摘要:

      背景 儿童期是认知功能发展的关键阶段，准确评估儿童的认知发展水平并了解其发展特点，对促进儿童健康成长具有重要意义。然而，传统认知评定方法通常依赖人工施测，存在效率偏低、趣味性不足等问题，较难适应儿童群体的评估需求，也难以在大规模应用中推广。目的 研发适用于儿童的数字化认知评估工具，为儿童认知功能的评定提供更便捷的测评手段。方法 基于经典心理学范式（Stroop任务、N-back、数字广度、空间定向和人脸命名范式），开发数字化认知评估工具。该工具包括颜色匹配、图形匹配、绿化家园、大收藏家和人脸与命名匹配五项任务，分别用于评定抑制控制、工作记忆、短时记忆、空间定向以及语义加工等核心认知功能。于2024年8月—2025年3月，纳入成都市某小学9~12岁学生，共750人。采用数字化认知评估工具进行评定。3个月后随机抽取其中的40名儿童，使用数字化认知评估工具及其对应的标准化心理学范式测试进行复测。采用Pearson相关分析，考查数字化认知评估工具的信度和效度：以数字化认知评估工具评分与对应标准化心理学范式测试评分的相关系数评估效度，以前后两次数字化认知评估工具评分的相关系数评估信度。比较不同年龄段、不同性别的儿童各项认知任务评分的差异。结果 共纳入有效样本699人。9~10岁为低龄组（n=460），11~12岁为高龄组（n=239）。男生356人（50.93%），女生343人（49.07%）。在信度分析中，数字化认知评估工具各任务前测评分与复测评分的Pearson相关系数为0.732~0.970（P均<0.01）。在效度分析中，数字化认知评估工具各任务评分与对应标准化心理学范式测试评分的Pearson相关系数为0.679~0.988（P均<0.01）。在颜色匹配任务中，年龄主效应和性别主效应均有统计学意义（F=31.071、21.198，P均<0.01）；在图形匹配任务中，年龄主效应、性别主效应以及年龄与性别的交互作用均有统计学意义（F=20.933、5.926、4.318，P<0.05或0.01）；在绿化家园任务中，年龄主效应有统计学意义（F=5.243，P=0.023）；在大收藏家任务中，年龄主效应有统计学意义（F=33.697，P<0.01）；在人脸与命名匹配任务中，性别主效应有统计学意义（F=27.016，P<0.01）。进一步分析显示，在五项认知任务中，女生组的高龄组评分均高于低龄组（P<0.05或0.01），男生组的低龄组颜色匹配和大收藏家任务评分低于高龄组（P<0.05或0.01）。在低龄组中，男生颜色匹配和图形匹配任务评分高于女生（P均<0.01）；在高龄组中，女生的人脸与命名匹配任务评分高于男生（P<0.01）。结论 本研究开发的数字化认知评估工具具有较好的信度和效度。9~12岁儿童的认知功能发展存在年龄与性别差异，不同认知维度的发展轨迹具有特异性。在低龄阶段，男生的抑制控制和工作记忆表现优于女生，但这一优势可能随年龄增长逐渐减弱；高龄阶段则转变为女生在语义加工能力方面更具优势。

英文摘要:

      Background Childhood represents a critical stage for cognitive development. Accurate assessment of children's cognitive abilities and understanding their developmental characteristics are essential for promoting healthy growth. However， traditional cognitive assessment methods typically rely on manual administration， presenting limitations such as low efficiency and insufficient engagement. These methods struggle to meet the assessment needs of children and are difficult to scale up for large-scale applications.Objective To develop a digital cognitive assessment tool for children， so as to provide a more convenient approach for evaluating children's cognitive functions.Methods Based on classic psychological paradigms （Stroop Task， N-back， digit span， spatial orientation， and face-name matching）， a digital cognitive assessment tool was developed. This tool includes five tasks including color matching， shape matching， greening the home， great collector， and face-name matching， designed to assess core cognitive functions such as inhibitory control， working memory， short-term memory， spatial orientation， and semantic processing， respectively. From August 2024 to March 2025， a total of 750 students aged 9–12 yeas old from a primary school in Chengdu were enrolled and assessed using the digital cognitive assessment tool. Three months later， 40 children were randomly selected for retesting using both the digital tool and its corresponding standardized psychological paradigms. Pearson correlation analysis was conducted to examine the correlation between the pre-test and retest scores of the digital cognitive assessment tool， as well as the correlation between the digital cognitive task scores and the corresponding psychological paradigm assessment results， in order to evaluate the reliability and validity of the digital cognitive assessment tool. Additionally， differences in scores across the cognitive tasks were compared among children of different age groups and genders.Results A total of 699 valid samples were included. The younger age group consisted of children aged 9–10 years old （n=460）， while the older age group comprised those aged 11–12 years old （n=239）. There were 356 boys （50.93%） and 343 girls （49.07%）. In the reliability analysis， the Pearson correlation coefficients between the pre-test and retest scores of each assessment task ranged from 0.732 to 0.970 （P<0.01）， indicating statistically significant results. In the validity analysis， the Pearson correlation coefficients between each task and its corresponding standard cognitive test ranged from 0.679 to 0.988 （P<0.01）. In the color-matching task， both the main effects of age and gender were statistically significant （F=31.071， 21.198， P<0.01）. In the shape-matching task， the main effects of age， gender， and their interaction were all statistically significant （F=20.933， 5.926， 4.318， P<0.05 or 0.01）. In the greening the home task， the main effect of age was significant （F=5.243， P=0.023）. In the great collector task， the main effect of age was significant （F=33.697， P<0.01）. In the face-name matching task， only the main effect of gender was significant （F=27.016， P<0.01）. Further analysis showed that within the female group， older group scored significantly higher than younger group in five tasks（P<0.05 or 0.01）. Within the male group， younger group scored lower than older group in both the color-matching and great collector tasks （P<0.05 or 0.01）. Within the younger group， boys scored significantly higher than girls in color-matching and shape-matching tasks （P<0.01）. In the older group， girls scored significantly higher than boys in face-name matching task （P<0.01）.Conclusion The digital cognitive assessment tool developed in this study demonstrates good reliability and validity. The development of cognitive functions in children aged 9–12 years old showed significant differences in age and gender， with specific developmental trajectories across different cognitive dimensions. At younger ages， boys outperformed girls in inhibitory control and working memory tasks， though this advantage diminished with age. At older ages， girls exhibited superior performance in semantic processing compared with boys.