Neurocognitive inefficiencies and eating disorder symptoms in college women

Data are for a study on neurocognitive factors and eating disorder symptoms in college women study. It was hypothesized that women with overlapping inefficiencies in set-shifting and global processing would have higher ED symptoms than those with standalone neurocognitive inefficiencies or no inefficiency. Data include ED symptom dimension scores (EDE-Q) as well as reaction time and error scores for the WCST (set-shifting) and Navon task (global processing). BIS scores were calculated to represent efficiency on the WCST and Navon task. This project contains the data file for all participants used in analyses as well as a codebook for the data. Three responses were missing completely at random. Missing values were imputed using the Expectation Maximization algorithm. Data were screened for univariate outliers using z scores (greater than +/- 3.29). Five univariate outliers were winsorized by replacing with the next within-bound value. Two multivariate outlier was identified using Mahalanobis distance and was removed from the sample, resulting in a final sample of 142 participants. Median splits of BIS efficiency scores were utilized to classify individuals into one of four neurocognitive profile groups: 1) overlapping (both), 2) central coherence (CC), 3) set-shifting (SS), and 4) no (neither) inefficiencies. Spearkman rank correlation analyses were run between eating disorder symptoms and BIS efficiency scores. Kruskal-Wallis tests were run to compare mean rank WCST and Navon task reaction time and error scores across neurocognitive profile groups. Differences in mean rank eating disorder symptom scores by neurocognitve profile were also tested using Kruskal-Wallis tests. Pairwise comparisons were performed using Dunn’s procedure with Bonferroni correction for multiple comparisons. Results showed that women with overlapping inefficiencies had higher eating disorder symptoms than those with no inefficiency.本数据集对应一项针对女大学生神经认知因素与进食障碍症状的研究。该研究提出假设：相较于仅存在单一神经认知功能缺陷或无神经认知功能缺陷的女性，同时存在任务转换（set-shifting）与整体加工（global processing）效率缺陷的女性，其进食障碍症状程度会更为严重。本数据集包含进食障碍症状维度评分（Eating Disorder Examination Questionnaire, EDE-Q）、威斯康星卡片分类测验（Wisconsin Card Sorting Test, WCST，对应任务转换任务）的反应时与错误率评分，以及纳冯任务（Navon task，对应整体加工任务）的反应时与错误率评分。本研究计算了行为抑制系统（Behavioral Inhibition System, BIS）评分，用以表征WCST与纳冯任务的完成效率。本项目包含所有用于分析的参与者数据文件，以及对应数据的编码手册。共有3条数据完全随机缺失，采用期望最大化（Expectation Maximization）算法对缺失值进行插补。通过z分数（阈值为±3.29）筛选单变量异常值，共5个单变量异常值通过替换为边界内的邻近值完成缩尾处理。采用马氏距离（Mahalanobis distance）识别出2个多变量异常值并将其从样本中移除，最终有效样本量为142名参与者。基于BIS效率评分的中位数分割法，将参与者划分为4类神经认知特征组：1）双缺陷组（同时存在两种缺陷）、2）中央连贯性缺陷组（central coherence, CC）、3）任务转换缺陷组（set-shifting, SS）、4）无缺陷组（均无缺陷）。本研究针对进食障碍症状与BIS效率评分开展斯皮尔曼（Spearman）等级相关分析；采用克鲁斯卡尔-沃利斯（Kruskal-Wallis）检验，对比不同神经认知特征组间WCST、纳冯任务的反应时与错误率评分的平均秩次；同时采用克鲁斯卡尔-沃利斯检验，分析不同神经认知特征组间进食障碍症状评分的平均秩次差异。采用邓恩（Dunn）法结合邦弗朗尼（Bonferroni）校正进行多重比较的两两事后检验。研究结果显示，同时存在两种神经认知缺陷的女性，其进食障碍症状程度显著高于无神经认知缺陷的女性。