Reconceptualizing language teaching: an in-service teacher education course in uzbekistan

Piloting. The best way to find out whether the created assessment 
tool works well is piloting. A test should not be administered without 
prior piloting. Such trialing can inform us about weaknesses the test 
might possess. Using a test without piloting may bring about serious 
consequences such as providing the wrong evidence about learners’ 
knowledge. Therefore, it is the department’s and particularly the test 
developers’ responsibility to strictly comply with the right procedures 
of the creation and use of assessments. The characteristics of the group 
which will be used for piloting a test should be similar to those of the 
target audience for the test itself. A teacher that piloted even one as-
sessment tool will realize how useful this can be. It can reveal the issues 
and the features of the test that the test developers otherwise might 
never be able to spot themselves. Thus, the evidence gathered from 
piloting is used to make the necessary changes and the whole process 
recommences. 
ACTION
Read the following reflection by an English language teacher who has learnt 
the main principles of language assessment and read extensively the literature. 
Please highlight the decisions and aims set by the teacher and pay attention to 
how each principle of language assessment is reflected.

---

118
RECONCEPTUALIZING LANGUAGE TEACHING
My teaching context is undergraduate students, which I taught the four 
skills of listening, reading, writing and speaking for Practical English. Apart 
from formative assessment that they have during the semester, they take one 
midterm and one final exam in each of the two terms in the academic year. 
Therefore, I usually deal with achievement tests based on the syllabus content. 
It means that my tests need to reflect what students have presumably learnt 
and assess their knowledge of the skills that they have developed through 
the course. In this endeavor, my concern has always been how to develop 
language tests used for classroom assessment in a professional way.
Having learnt the significance of different test qualities, I find it important 
to bear in mind that “a test cannot be valid for more than one purpose” (Hen-
ning, 1987, p. 89), which I used to ignore without realizing how inappropriate 
a certain test can be. I made no serious attempt to check the validity and re-
liability of the tests that I created or selected for assessment. All my decisions 
have been based on my own intuitions and the instructions from the admin-
istration and the head of my department. However, what I am inclined to do 
as of now is to make validity and reliability central in the development of my 
assessment tools. My aim from this is to make sure that students are tested 
for what they are intended to be tested and the tasks across all the variants of 
tests are of equal difficulty. 
Usually my department administers midterm and final tests in several 
slots within the same day by grouping students, as there are over 400 students 
to be tested. As a result, my task as a test developer is to create several variants 
for every slot. At the same time, I have to make sure that every variant is no 
different from another in terms of difficulty and other characteristics. Here 
my responsibility is to construct equivalent tests, or what Weir calls “parallel 
forms” (2005, p. 250) so no matter what variant is given to a certain student, 
he/she should be able to perform with the same results, which is the concern 
of reliability. An important consideration is checking a testing instrument by 
piloting. I have been through experiences when my test turned out to be too 
easy or too difficult. I was usually given feedback from other teachers in my de-