【作者】 陈红斌；

【导师】 凌文辁；

【作者基本信息】 暨南大学 ， 应用心理学， 2004， 硕士

【摘要】 情景判断测验技术(Situational Judgement Tests，SJT)是近来发展较快的一种职业能力评价技术。以其贴近工作实践，更高的表面效度，更强的针对性，较低的开发和使用成本，而受到国外众多学者和企业的重视，被广泛地研究和应用。但是在国内，真正严格意义上的情景判断测验的研究与应用还十分有限，因此，本研究在“情景判断测验技术理论”的指导下，以人寿保险代理人为对象构建了“人寿保险代理人销售能力情景判断测验”，以期在作理论探索的同时开发出对实践工作有所帮助的评价工具。 本研究由八个阶段的工作组成：(1)采用结构化的访谈、查阅文献和现场观察，对人寿保险代理人的工作进行分析；(2)在工作分析结果的指导下，采用关键事件访谈、查阅文献资料收集典型情景；(3)采用德尔斐法评价所收集到情景的典型性水平，剔除、修改不合格的情景；(4)使用开放式问卷收集各种情境下的可能反应；(5)用德尔斐法对各种反应的典型性加以评价，剔除、修改不合格的反应；(6)预试，依据测验的项目区分度、信度和效度评价结果修改测验；(7)正式施测，确定测验项目区分度、信度和效度的水平；(8)编写测验说明书。 研究采用克隆巴赫(Crombach α)一致性系数和分半信度评价测验的信度水平；以《保险推销员(职业)自我效能感量表》、被试将自己截至受访时前三个月的月平均绩效与同事的比较后确定的水平，以及被试将自己截至受访时前三个月的日平均拜访量与同事的比较后确定的水平作为效标，来评价测验的效度水平。 本研究正式施测结果显示，由28个项目组成的测验的一致性系数为0.863，分半信度系数为0.822，达到了心理测量学的要求；测验与自我效能感、工作绩效和拜访客户量的相关分别为0.368、0.353和0.297，均达到了.01的显著性水平，且接近国外同类测验效度的平均水平(0.34)。 笔者希望通过这一研究，在方法上借鉴国外情景判断测验开发模式的基础上，探索新的开发模式；在实践方面，开发出“人寿保险代理人销售能力情境判断测验”，对现实工作起一定的帮助和推动作用。更多还原

【Abstract】 Situational Judgement Tests (SJT) is one of most quickly promoted evaluation skills lately. Because it is close to work related contexts, has high face validity, and the cost for promoting and use is comparatively low, so it has been used in a very large field.In this research, under the direction of SJT, the author construct a test focus on sales ability of insurance agent.The whole construct process is build up with eight steps. First, doing job analysis by structure interview, consulting literature and field observe. Second, collecting representative situation by the guide of job analysis result, through interview and consulting literature. Third, using Delphi method evaluate the representative level of situations. Forth, using open questionnaire to collect reactions. Fifths, evaluate the representative level of reactions by Delphi method. Sixth, pre-test, analysis item differential validity, test reliability and validity, then amend the test. Seventh, testing, make sure item differential validity, test reliability and validity. Eighth, write uservs guide.This research use Crombach a and split half reliability coefficient to evaluate the reliability level of the test. The criterionrelated validities are the relation coefficient between test score and "The scale for self- efficiency feeling of insurance agent", object’s job performance in latest three month, the number of client object visit in latest three month.The test is build up with 28 items. It’s Crombach a is 0.863, split half reliability coefficient is 0.822. The correlation coefficient between test score and "The scale for self- efficiency feeling of insurance agent", object’s job performance in latest three month, the number of client object visit in latest three month is 0.386,0.353,0.297.Through those effort the author want to explore a new construct model on the base of oversea model in academic field, and facilitate daily work in practice field, on the other hand.更多还原