Abstract: Background　Anxiety is one of the most common mental problems nowadays，and wrong cognition can lead to anxiety.Therefore，it has become a hotspot of psychology research to alleviate anxiety through cognitive bias modification.The anxiety in military personnel is very important，while few domestic researches pay much attention to military cognitive bias，especially the research on interpretation bias.Objective　To explore differences of interpretation bias among officers and soldiers with high and low levels of anxiety，and investigate whether there is universality across self /non-self relevant situations.Methods　Convenient sampling was applied in the study，selecting 118 military staff who consented to fill out State-Trait Anxiety Inventory （STAI） from August to September in 2018．Those who scored higher by one standard deviation were taken as the high anxiety group（SAI>46，n=24），while those lower by one standard deviation were taken as the low anxiety group（SAI<28，n=21）．Self-made ambiguous scenarios paradigm were used to finish the experiment.The 3-factor mixed experimental design of 2 was used．（groups：high anxiety group & low anxiety group） ×2（related type：self relevant situation & non-self relevant situation） ×2（interpretation titer：positive interpretation & negative interpretation），with groups as inter-subject variable，related type and interpretation titer as inner-subject variables.Scores on different interpretations were used as dependent variables.Results　Interaction of groups and interpretation titer showed significant differences〔F（1，43）=14.290，P<0.01〕，while there was no significant differences between related type and groups〔F（1，43）=0.034，P>0.10〕，related type and interpretation titer〔F（1，43）=0.065，P>0.10〕，and interaction among groups，related type and interpretation titer〔F（1，43）=0.534，P>0.10〕.Simple effect analysis of groups showed significant differences in both positive interpretation〔F（1，43）=8.410，P<0.01〕 and negative interpretation〔F（1，43）=13.380，P<0.01〕.High anxiety group scored lower than low anxiety group in positive interpretation（P<0.05），while higher in negative interpretation（P<0.05）．Conclusion　Compared with low anxiety group，high anxiety group showed no significant difference in interpretation bias under different self-relevant situations.In ambiguous scenarios，officers and soldiers with high level of anxiety scored lower in positive interpretation while higher in negative interpretation than those in the low anxiety group.

Key words: Anxiety, Military personnel, Interpretation bias, Ambiguous scenarios test, Transactional analysis, Positive interpretation, Negative interpretation

摘要： 背景　焦虑是当今最常见的心理问题之一，不合理认知导致焦虑，因而通过认知偏差矫正来缓解焦虑已成为心理学界的研究热点。军人焦虑问题更不容小觑，而国内关于军人认知偏差的研究还比较薄弱，解释偏差的研究也更少。目的　探讨不同状态焦虑军人解释偏向的差异，并考察解释偏向是否存在跨自我/他人相关的普遍性。方法　采用方便抽样法，于2018年8—9月选取中国人民解放军某部队118名军人，对其施测《状态-特质焦虑问卷》（STAI），取状态焦虑问卷（SAI）得分大于该样本均值的1个标准差为高状态焦虑组（>46分，n=24），小于均值的1个标准差为低状态焦虑组（<28分，n=21）。自行编制军事模糊情境句，采用模糊情境试验范式完成行为学试验。采用2（组别：高状态焦虑组和低状态焦虑组）×2（相关类型：自我相关和非自我相关）×2（情绪效价：积极解释和消极解释）三因素混合试验设计，其中组别为被试间变量，相关类型和情绪效价为被试内变量。因变量为被试在不同解释上的得分。结果　组别和情绪效价的交互作用显著〔F（1，43）=14.290，P<0.01〕。组别与相关类型〔F（1，43）=0.034，P>0.10〕、相关类型与情绪效价〔F（1，43）=0.065，P>0.10〕的二重交互作用，以及组别、相关类型和情绪效价的三重交互作用〔F（1，43）=0.534，P>0.10〕均不显著。焦虑分组在积极解释上的简单效应显著〔F（1，43）=8.410，P<0.01〕，高状态焦虑组在积极解释上的得分低于低状态焦虑组（P<0.05）。焦虑分组在消极解释上的简单效应显著〔F（1，43）=13.380，P<0.01〕，高状态焦虑组在消极解释上的得分高于低状态焦虑组（P<0.05）。结论　相较低状态焦虑军人，高状态焦虑的军人在不同自我相关条件下的解释偏向差异不显著，但在面对模糊情境时，高状态焦虑军人的积极解释得分更低，消极解释得分更高。

关键词: 焦虑, 军事人员, 解释偏向, 模糊情境试验, 交互分析, 积极解释, 消极解释