Feedback on their performance

considered a minimal passing grade). 
Construct-Related Evidence 
A third kind of evidence that can support validity, but one that does not play as large a role 

---

 
 
98 
classroom teachers, is construct-related validity, commonly referred to as construct 
validity. A construct is any theory, hypothesis, or model that attempts to explain observed 
phenomena in our universe of perceptions. Constructs may or may not be directly or 
empirically measured-their verification often requires inferential data. 
D. Authenticity 
An fourth major principle of language testing is authenticity, a concept that is a little 
slippery to define, especially within the art and science of evaluating and designing tests. 
Bachman and Palmer (1996, p. 23) define authenticity as “the degree of correspondence of 
the characteristics of a given language test task to the features of a target language task,” 
and then suggest an agenda for identifying those target language tasks and for 
transforming them into valid test items. 
E. Washback 
A facet of consequential validity, discussed above, is “the effect of testing on teaching and 
learning” (Hughes, 2003, p. 1), otherwise known among language-testing specialists as 
washback. In large-scale assessment, wasback generally refers to the effects the test have 
on instruction in terms of how students prepare for the test. “Cram” courses and “teaching 
to the test” are examples of such washback. Another form of washback that occurs more in 
classroom assessment is the information that “washes back” to students in the form of 
useful diagnoses of strengths and weaknesses. Washback also includes the effects of an 
assessment on teaching and learning prior to the assessment itself, that is, on preparation 
for the assessment. 
 
F. Applying Principles to the Evaluation of Classroom Tests 
The five principles of practicality, reliability, validity, authenticity, and washback go a long 
way toward providing useful guidelines for both evaluating an existing assessment 
procedure and designing one on your own. Quizzes, tests, final exams, and standardized 
proficiency tests can all be scrutinized through these five lenses. 
Are the test procedures practical? 
Is the test reliable? 
Does the procedure demonstrate content validity? 
Is the procedures face valid and “biased for best”? 
Are the test tasks as authentic as possible? 
Does the test other beneficial washback to the learner?