In most human situations we are best tested in something when we are required to perform a sampling of the criterion behavior. But there are a few cases of highly specialized an~ sophisticated testing instruments which do not have high content validity yet are nevertheless valid. Projective personality tests are a prime example. The Thematic Apperception Test and the Rorschach "inkblot" tests have little content validity, yet they have been shown to be accurate in assessing certain types of deviant personality behavior. Other well-known psychological tests have little content validity. The Micro-Momentary Expression test (MME) is a test of empathy that requires subjects to detect facial changes in a participant in a conversation. The more facial changes a testee detects, the more empathic he or she is said to be. Such a test has little con- tent validity, especially if the astute detection of facial changes might be argued to require field independence, which has been shown to correlate negatively with empathy! A test of field independence as a prediction of language success in the classroom is another example of a test with potentially good criterion validity but poor content validity in that the ability to detect an embedded geometric figure bears little direct resemblance to the ability to speak and hear a language.

Construct Validity
A second category of validity that teachers must be aware of in considering language tests is construct validity. One way to look at construct validity is to ask the question: does this test actually tap into the theoretical construct as it has been defined? "Proficiency" is a construct. "Communicative competence" is a construct. "Self-esteem" is a construct. Virtually every theoretical category we have discussed in this book is a theoretical construct. Tests are, in a manner of speaking, operational definitions of such constructs, in that they operationalize the entity that is being measured (see Davidson, Hudson, and Lynch 1985). A teacher, then, needs to be satisfied that a particular test is an adequate definition of a construct. A general proficiency test that consists of, say, grammatical judgment items, reading comprehension items, and listening comprehension items is defining "proficiency" as either consisting of, or being correlated with, those three modes of performance.

In many cases such theoretical constructs are perceived as being adequately defined in the content of the test its elf. But when there is low, or questionable, content validity in a test, it becomes very important for a teacher to be assured of its construct validity. In this instance, validation of the construct has to be empirically demonstrated by means of research that shows that the behavior required of the testee is correlated with the total construct of behaviors in question. For example, the empirical justification for using the MME as a test of empathy is found in research that shows the MME to be correlated with other tests of empathy. The Embedded Figures Test of field independence, in which the testee is to discern smaller geometric shapes within larger and more complex geometric designs, has likewise been related in research studies to other forms of assessing field independence. If you were to claim that such a test is valid for, say, predicting success in a second language, you would be forced to do (or find) research that would empirically demonstrate the correlation of scores on the Embedded Figures Test with scores on other measures of language aptitude. The construct underlying such a claim would theorize that the game cognitive strategies or styles required to perform well on the Embedded Figures Test are also required for successful learning of a second language.

Validity is a complex concept. However, it is indispensable to the teacher's understanding of what makes a "good" test. If in your language teaching you can attend to the practicality, reliability, and validity of tests of language, whether those tests are classroom tests related to a part of a lesson or final exams or proficiency tests, then you are well on the way to making accurate and viable judgments about the competence of the learners with whom with whom you are working.

Kinds of Tests
There are many kinds of tests, each with a specific purpose, a particular criterion to be measured. The purpose of this chapter is not to expound on the many varieties of tests, nor to instruct you on how exactly to devise even a few varieties. However, in examining the general underlying principles of language testing, it is appropriate and necessary to devote a few brief words to outlining some categories of language tests. Your training as a teacher of a particular language should then involve the more specific matters of test construction and interpretation for that language.