在迫选测验研究领域，有学者认为被试在作答迫选测验时，如果在两个选项上的真 实得分有较大差异，那么他们就能比较容易地选出符合自己的一项，决策的过程也比较 快。如果被试在两个选项上的真实得分较为接近，则会导致相反的表现，可能更容易受 到其他无关因素的影响。但相关假设尚缺乏实证研究的支持，不确定被试在作答一个迫 选测验时的决策过程是否真正受到选项真实得分之间差异性的影响。在应聘场景中被试 有较高的作假动机时，其决策过程是否也受到选项真实得分之间差异性的影响，以及是 否会受到选项间社会称许性的影响而做出不一样的选择，这些尚未有研究者来探讨。

本研究拟通过设计真实作答场景和作假作答场景，诱发被试对迫选测验的诚实作答 和作假作答反应，采集作答结果、反应时和眼动指标数据来分析被试的认知决策过程， 探索迫选测验的选项间的匹配关系对作答结果的影响。选项配对分为选项真实得分差异 匹配和接近匹配两个水平，其中差异匹配是指被试在两个选项上的得分存在差异，接近 匹配是指在两个选项上的得分是相同的。选项上的真实得分是提前让被试作答Likert式大 五人格量表而获得，由于量表上的作答结果各不相同，因而为每个被试匹配生成的迫选 测验也因人而异。

结果发现在诚实情景下，对于真实得分差异匹配的迫选题对，被试倾向选择真实得 分更高的一项，也表现出对高真实水平选项更高的注视与扫视水平。在作假情景下，被 试未表现出上述的作答与眼动倾向，反而更趋于选择具有更高称许性的选项，但未在高 称许性选项上表现出更高的注视和扫视水平。同时，在诚实情景下，对于真实得分接近 匹配的迫选题对，被试表现出相较真实得分差异匹配的迫选题对更高的作答用时及注视 与扫视水平，而且在作答真实得分接近匹配题对时，其在高称许性选项上表现了更高的 注视与扫视水平。

研究表明，被试在迫选式人格测验上的作答会受到题目匹配关系的影响。当迫选题 目中两个选项的真实分数较为接近时，被试会面临作答决策的困难。如果真实回答其可 能会随机选择一个选项，还会关注题目中与自身无关的因素，如在高称许性选项上表现 出更高的注视和扫视水平。如果被试有作假回答的动机则更容易选择高称许性的选项。 因而，编制迫选测验时，应尽量减少或控制选项间的真实水平接近匹配。此外，社会称 许性的匹配仍然非常重要，用于高利害选拔场景的迫选测验应尽可能对选项称许性进行 精确匹配，控制被试的称许性作答倾向，提高迫选测验施测有效性。

Researchers believe that when answering the forced choice test, if there is a significant difference in the true scores of the two options, the subjects can easily choose the one that suits themselves, and the decision is also faster. If the true scores of the two options are relatively close, it will make it difficult for the subjects to determine which one is more suitable, and the decision-making will be longer, which may be more susceptible to the influence of other unrelated factors. However, this hypothesis still lacks empirical research support, and it is uncertain whether the decision-making process of participants when answering a forced choice test is truly influenced by the differences between the true scores of the options. When participants have a high level of fraudulent motivation in the job application scenario, whether the decision-making of subjects is still influenced by the differences between the true scores of the options, and whether they will make different choices due to the influence of social desirability between the options, has not yet been explored. This research aims to induce subject' honest and fake responses to forced choice tests by designing real and fake response scenarios, collect response results, response time, and eye tracking indicator data to analyze subject' cognitive decision-making process, and explore the impact of matching relationships between the options of forced choice tests on response results. The block matching is divided into two levels: difference matching and close matching base on true score. Difference matching refers to the differential in scores between the subjects on two options, while close matching refers to the same scores on both options. The true score of the subjects on the options was obtained by Likert scale. Due to the different responses between subjects, the forced choice tests generated for each participant's matching also unique. The research showed that: In honest scenario, when there were differences between the forced-choice options compared to true score, subjects tended to choose the option with a higher score, and also showed higher levels of fixation and saccade in these options. In fake scenario, the subjects did not exhibit the aforementioned tendencies towards answering and eye tracking data, they tended to choose options with higher social desirability, but did not exhibit higher levels of fixation and saccade on those options. In honest scenario, when subjects answering items that closely matched with true score, they exhibited longer response time and higher levels of fixation and saccade compared to items match by another method. Moreover, the subjects showed higher gaze and saccade levels on the highly social desirability options. In fake scenario, there was no difference in the answering time and eye tracking indicators between the two matching methods. When subjects answering items that closely with social desirability, the participants did not show a tendency to choose options with higher social desirability. This research has shown that the matching of forced-choice items will affect individuals' responses. When the true score corresponding to the two options in forced-choice items are relatively close, it will bring additional difficulties to the answering process, which may cause individuals to select an option randomly, and showed higher gaze and saccade levels on the highly social desirability options. If the subject has a motivation to faking, they are more likely to choose the option with higher social desirability. In the future development of forced-choice scales, this phenomenon should be minimized or controlled as much as possible. In addition, the matching of social desirability is still important, and the forced-choice tests used in highrisk selection scenarios should accurately match the social desirability as much as possible, control the respondents' tendency to faking, and improve the effectiveness of forced-choice tests.