基于视线追踪的军校心理测量应答倾向性识别与验证

Recognition and validation of response tendency of military psychological measurement based on eye-tracking

目的采用眼动与机器学习技术预测被试回答量表项目的真实倾向性,从而规避量表答题中的掩饰性问题。方法采用眼动技术同步采集被试量表答题过程中的眼动数据,通过支持向量机(SVM)与主成分分析(PCA)相结合的机器学习算法对被试答题过程中眼动指标进行分类,预测被试回答量表项目的真实倾向性。结果采用结合PCA与径向基核函数的SVM算法对180个眼动指标特征进行分类,预测被试回答量表项目的真实倾向性平均准确率达到91.8%,可有效实现被试真实倾向性的客观预测。结论眼动和机器学习技术可有效实现被试答题真实倾向性的客观预测,能显著提高心理检测的准确性,验证了多质融合心理选拔技术的有效性。本研究为后续多种认知神经技术融合对量表项目应答倾向性预测提供了思路和参考,为规避心理测量中掩饰性提供了新途径。

Objective To predict the true tendency of the subjects as answering scale items by eye movement and machine learning technology so as to avoid disguised questions in scale.Methods The eye movement technology was used to synchronously collect eye movement data of subjects in the process of answering questions.The machine learning algorithm combined with support vector machine(SVM)and principal component analysis(PCA)was used to classify the subjects eye movement indicators in the process of answering questions so as to predict the true tendency of the subjects when answering the scale items.Results SVM algorithm combined with PCA and radial basis kernel function was used to classify 180 eye movement index features,and the average accuracy rate of predicting the true tendency of the subjects answering scale items reached 91.8%,which could effectively predict the true tendency of the subjects.Conclusion Eye movement and machine learning technology can effectively and objectively predict the true tendency of the subjects as answering scale questions.They can significantly improve the accuracy of psychological detection,and verify the effectiveness of multimodal fusion psychological selection technology.This study provides new insights and references for the subsequent fusion of multiple cognitive neurotechniques to predict the response tendency of scale items as well as a new way to avoid the disguise in psychological measurement.