師資生潛能測驗組合：偽好作答效果之檢驗

Previous research has shown that when academic competence is used as thesole criterion of student teacher selection, it may be a poor indicator of howcandidates perform when they become in-service teachers (Wright, 2015). Thus,in the future teacher training institutions will need to identify alternative indicatorsfor determining which students to accept to teacher training programs. The Testsfor Student Teacher Potential that we previously developed incorporate toolscommonly used in the job market, such as situational judgement tests, workvalues scales, and personality scales. We spent several years compiling testsand building online test systems. Three systems providing more comprehensiveindicators for teacher training institutions have gone live: Situational JudgmentTests for Teachers (Chao, Sung, & Haung, 2020), Work Values Scale forTeachers (Chao, Sung, & Hsu, 2018), and Personality Scale for Teachers (Chao& Sung, 2020). However, in the context of selection, non-cognitive tests may beaffected by factors such as response bias and social desirability, resulting in testscores that fail to reflect the test-taker’s actual affective state (i.e. “faking good”;McFarland & Ryan, 2006). As applicants, students may intentionally orunintentionally inflate their self-assessment, leading to self-deception andimpression management in tests (Becker & Colquitt, 1992). Recent studies havefound that “faking good” threatens the measurement invariance of personalitytests, distorts a respondent’s ranking order, and affects subsequent selectiondecisions (Krammer, Sommer, & Arendasy, 2017). During the two years of itsduration, this research project will use an experiment and review of test data fromactual selections to determine whether applicants are “faking good” in the Testsfor Student Teacher Potential. The experiment will divide respondents into fourgroups, each receiving a different pre-test instruction. The experimental goal is todetermine whether students fake when they take the Tests for Student TeacherPotential and to identify which of thethree tests that have gone live is more susceptible to faking behavior. To gathertest data on actual selections, some teacher training institutions will requirestudents to complete the Tests for Student Teacher Potential, and we shall alsodesign a questionnaire asking students about the strength of their motivation toenroll in a teacher training program. We will then examine whether there is anycorrelation between the stated degree of motivation and their test results. Last, bycombining these two studies, the project will explore ways to prevent studentfaking on the Tests for Student Teacher Potential. In addition to achieving a morein-depth academic understanding of the characteristics of each of the three tests,the results of the studies will help improve current test tools by designingmechanisms to minimize faking.