This mixed-methods study sought to examine any significant impact of Iranian English as a Foreign Language (EFL) teachers' collaboration in rating speaking tests and brainstorming theoretical concepts of language assessment (LA) on their literacy, perceptions, and practices of speaking assessment. Forty-one Iranian EFL teachers selected through purposeful sampling were asked to fill out a questionnaire, including 20 components of LA, so as to measure their knowledge and perceptions of LA components. Then, 21 of the teachers volunteered to score the video-projected speaking performance of 12 international students in the Key English Test (KET) using KET speaking rating scales. Having received six in-service speaking assessment training sessions based on their collaboration, the 21 participating teachers completed the same questionnaire and scored the same videos again. An ANCOVA was used to compare the quantitative data from the questionnaire and teachers’ scoring of KET speaking tests before and after the treatment and also to investigate any significant differences between the novice and experienced Iranian EFL teachers in terms of their knowledge and perception of LA, and the assessment of KET speaking tests. Moreover, the 21 volunteered teachers were interviewed to further discover their perceptions about the in-service speaking assessment training course they attended. The findings revealed that the collaboration of the Iranian EFL teachers in rating speaking tests and brainstorming theoretical concepts of LA significantly impacted their knowledge, perceptions, and practices of LA. However, the effects were unequal and different for novice and experienced EFL teachers.
