Evaluating Assessments

This week’s readings centered on assessing students in language classrooms.  The most important thing that I learned from Shohamy’s article is that teachers need to choose tests that match the construct they are trying to measure.  Although Shohamy didn’t talk in depth about the concept of constructs, she discussed how things such as testers’ background knowledge impacting test scores (204).  Another topic she discussed was the effect of different test types on student results.  A student’s high performance on a multiple choice test could reflect natural test-taking ability more so than language proficiency.  Furthermore, it is difficult to assess a student’s conversational skills if they are talking into a tape recorder rather than with a live human being.  Teachers need to consider their intended construct and what they want to learn from the test before creating tests and test items.

While Shohamy narrowed her topic to discourse and testing, Brown discussed many important principles in relation to language and assessment.  He analyzed the differences between assessment and testing, characteristics of good tests such as validity and reliability, and how washback impacts teaching and learning.  We have learned about many of these topics already in our assessment class with Dr. Kang, so these chapters were basically review for me.  However, I was still intrigued by some interesting ideas. The most interesting part of chapter 23 to me was the question about whether or not teachers should always be assessing students (445).  Although this is a very complex question, I think that yes, teachers should always be assessing their students.  Not all of this assessment is formal, but teachers need to always think about how they can help students, what errors students are producing, and take note of how students are performing.  To me, if an activity doesn’t alert teachers or students themselves about how they are doing, then it isn’t a strong activity.  When students are working independently and in groups, the activity should in some way monitor the progress of students for both teachers and students themselves.  Informal class activities can be so valuable because they can serve as a formative assessment that can be used to improve both teaching and learning. 

Chapter 24 covers more items to consider when developing classroom tests.  Through the little experience I have, I have already begun developing my approach to testing.  Firstly, I believe criterion-referenced tests are much better suited for individual classrooms than norm-referenced tests.  Students should be graded on standard criteria and what they can and cannot do with the language; the performance of their peers shouldn’t affect their grades.  I am also in favor of an approach that utilizes performance-based assessment.  As teachers we need to find out what students can do with the language in authentic settings.  If students cannot apply what they have learned to authentic tasks, then our class is not practical and will not help them in the real world.  In place of essays, students could write e-mails to Spanish speakers, journal articles for native speakers, and write personal statements for job interviews.  21^st century education calls for students to be able to use the skills they learn in school in the workplace.  Traditional multiple choice tests are no longer adequate for evaluating the skills of our students.