February, 8th 2010
What I have learned for the first four weeks of the course has prompted me to initiate new policies in testing administration together with training workshops for teachers in my division. The aims are to work toward better quality test practice: using tests for reflection on students’ progress rather than failing or passing students and improving the present test giving practice in weak areas such as validity, reliability, wash back. My reflection on the three testing models is that we should move from the model with mid-term plus final assessment to the model where the assessment process is parallel to the teaching/learning process through the course. To achieve this aim new policies should emphasize the assessment component in the professional activity package for novice teachers to qualify test developers. The assessment plan is expected to empower the teaching staff with relevant knowledge and skills so that they can integrate assessment into learning process and improve the quality of all assessment. Once sufficient training has been provided more diagnostic and formative assessment will be introduced to the instructional program. More alternative forms e.g. portfolio, surveys will be employed as these assessments ask students to perform, create, produce or do something; thus tap more sophisticated thinking and problem solving skills (Herman, Aschbacher and Winters, 1992). Instructors also need to train students to enable them to initiate self-assessment when required. It must be pointed out that making assessment one integral part of learning will be a long-term aim as alternative assessment requires new instructional and assessment roles for teachers (Herman, Aschbacher and Winters, 1992), which suggests a high level of training. To improve test quality workshops given by experts from other schools as well as experienced teaching staff should be organized. They should be practical in nature with the purpose of developing good tests from available sources. Besides training they, together with trainees, could help by finalizing test developing checklists. For example, assessments should be meaningful tasks, related to personal objectives and contextualized. To secure more reliability, marking criteria, especially for writing and speaking, should be double-checked by experienced test developers and administration procedures between classes need to be better standardized (Coombe, Folse & Hubley, 2007). Discussion as well as moderation should precede the marking sessions for productive skills. Listening sections which involve language production such as dictation and re-production should call for discussion and moderation too. Additionally, validity is the area that needs more attention especially in listening and reading, where it is more difficult to specify students’ characteristics and come up with the right specifications. For writing, topics and contexts should be given more work as they are easily affected by cultural bias. It would further help if procedures in developing tests were elaborated with clear instructions about each stage, e.g. specifications must be consulted before the test development process (Coombe, Folse & Hubley, 2007). Other problems in the present test making practice as revealed in my test critique such as inadequate care given to the issues of background knowledge, skill contamination, backwash (Coombe, Folse & Hubley, 2007) should be all given due care. The suggested policies and training will definitely benefit the students. As new assessment is aimed at identifying students’ progress the instructors will know the students better and become more effective in assisting them. For example, teachers can plan more effective lessons and more regular feedback will be provided to the learners to direct their learning. Effective feedback procedure will make tests a critical link in teaching (Angelo & Cross, 1993). Classroom...