With the national education focus on accountability for student achievement, teacher education programs must select and train teachers who can impact student learning. The problem investigated in this study was the lack of congruence between the traditional methods of selecting teacher candidates and research findings on teacher effectiveness. The literature on teacher effectiveness displays the importance of teacher dispositions, while the selection criteria for teacher education programs includes quantitative data (such as standardized test scores and GPA) that do not significantly predict teacher performance.

Two instruments designed to predict teacher effectiveness, the TeacherInsight™ Interview and the Teacher Work Sample (TWS), were examined in this study. The purpose was to investigate the relationship between TWS scores and TeacherInsight™ Interview scores and to examine any relationship of demographic or academic variables on these scores. Study subjects were teacher candidates enrolled in the capstone student teaching classes in the College of Education and Behavioral Sciences (CEBS) at Western Kentucky University in the fall 2007 and spring 2008 semesters (N = 396).

Three research questions guided the study: (1) Does teacher candidate performance on the Teacher Work Sample predict scores on the TeacherInsight™ Interview? (2) Are demographic and/or academic variables related to scores on the Teacher Work Sample? (3) Are demographic and/or academic variables related to scores on the TeacherInsight™ Interview?

These predictor variables were analyzed for their relationship to TeacherInsight™ Interview score: (a) TWS score, (b) gender, (c) age range, (d) ethnicity, (e) certification, (f) grade point average, (g) ACT score, and (h) Praxis II: PLT score.

Multiple regression was used to answer research questions one and three. TWS scores did not predict performance on the TeacherInsight™ Interview. Only one variable, certification level, had a statistically significant relationship with TeacherInsight™ Interview score. For research questions two and three, ANOVA, Pearson correlations, and an independent t test were used to determine if demographic (age range, ethnicity, certification level, and gender) or academic (ACT, GPA, and Praxis II: PLT score) variables related to scores on TWS. Three variables (gender, Praxis II: PLT scores, and certification level) had statistically significant relationships with both teacher assessments. Also, ACT had a significant correlation with TWS score.