The Impact of Rater Bias on the Language Performance Assessment Scores of Iranian Foreign Language Teacher Candidates

Utilizing the scores obtained from a teacher entrance test used in an English language institute as a means of selection, the researchers selected 100 out of 121 female teacher candidates to participate in this study. Furthermore, a reading, writing, and listening test was administered to the candidates to exclude those candidates with low and high proficiency. Based on the results obtained from the tests, the number of participants decreased to 30 and they were requested to come in for oral interviews; they were those who were interviewed twice by two different groups of male and female raters. The results analyzed through correlational analysis and descriptive statistics indicated that the interview scores of the teacher candidates, as measured by the first group of female raters () was in high correlation with those in class performance (p). However, there was no correlation both between the interview scores of the female teacher candidates (), assessed by the male raters and their in class performance (P) and the () - () pair. It can be concluded that rater bias might have had an effect on the Iranian female teacher candidate s test scores. The subjects were also divided into attractive and unattractive groups and further assessed by the fourth and fifth group of female and male raters to indicate whether female sex appeal affects the test scores or not. The results showed that the mean differences between the AF-AM (attractive female-attractive male) pairs were significant although the mean differences between the NAF-NAM (nonattractive female-non-attractive male) (P=.05, t=1.131) were not significant.

Keywords:

Correlational analysis , Oral Interview , Rater Bias , t-test