مجید نوروزی

Grades represent one of the most common sources of evidence of student achievement in classrooms, though their relationship with test scores has remained understudied, particularly in settings such as in Iran, where English is taught as a foreign language. The purpose of this study was to investigate the relationship between graded and tested achievement with respect to gender and proficiency level differences. Teacher-assigned grades and standardized achievement test scores of 693 Iranian learners of English taught by 15 teachers were examined. Primary analyses focused on the validity of teacher grades and the subsequent Pearson correlation coefficients revealed that grades associated positively with externally-validated test scores obtained from reliable tests, an indication of the validity of teacher grading. Additionally, the results of independent-samples t-tests showed that female students outperformed male students on achievement tests, but with fluctuations across proficiency levels. Higher proficiency levels gave male participants an advantage over female participants in achievement tests. Moreover, male teachers were found to grade female participants more accurately than their female counterparts. Implications are discussed for informing teachers about the validation of their grading practices, as well as for teacher education programs and teachers’ professional development.

Although grades are the most ubiquitous currency of educational measurement around the globe, their meaning, particularly in understudied settings as in Iran, is still shrouded in mystery. The purpose of this study was to investigate EFL teachers’ class grades by focusing on the less verbalized aspects of grading to see what a grade means. Five Iranian English language teachers working part-time in a private EFL institute were interviewed using the repertory grid interviewing technique, Kelly’s (1955) unique data collection instrument used extensively in personal construct theory (PCT). The results of the content analysis revealed that of the 92 elicited constructs, over 70% were categorized as non-academic, pointing to a heavy reliance on such criteria for grading, and consequently leading to the invalidity of assigned grades. Further, the results of principal component analysis (PCA) of each teacher’s elicited constructs endorsed hodgepodge grading by demonstrating single main components that accounted for the most variation in teacher grading and that comprised both academic and non-academic factors. However, this phenomenon was interpreted slightly differently when seen from the PCT perspective. Implications of this study for teacher professional development and teacher education programs are discussed.