6 Assessing productive and interactive skills Personal refl ection

 Part 
II
Trying out material
With more open-ended and extended performance assessments, it is 
particularly helpful to obtain evidence of the kinds of response that 
a task prompt actually generates (Association of Language Testers in 
Europe/Council of Europe, 2011). A fi rst step is for writers to 
attempt the task themselves and for colleagues to do the same. This 
can also help the scoring process because these responses can be used 
as benchmarks for expert performance to assist scorers. However, 
all assessment material should be tried out with groups of learners 
who are as similar as possible to the assessees who will take the 
operational assessment. 
In addition to analysing responses to see that the assessees were able 
to produce the kinds of language intended, the designers have the 
opportunity to fi nd out from the assessees and examiners whether they 
experienced any diffi culties in following the tasks or identifi ed any 
other potential problems with the material.
A complicating factor in face-to-face assessments of speaking is that 
the assessor actually delivers the assessment. Assessor training and 
monitoring must therefore also be part of the quality control process. 
Are assessors able to adhere to their instructions? What impact do 
these instructions have upon the kind of performance that is elicited? 
In formative assessment, what techniques for supporting or scaffolding 
learner performance are most effective? What follow-up activities 
might be used to challenge learners to extend their learning? The 
material and procedures should be reviewed and accepted, revised or 
rejected as appropriate.