What is a Test?
Whether we realize it or not, we test every day in virtually every cognitive effort we make. When we read a book, listen to the news on TV, or prepare a meal, we are testing hypotheses and making judgments. Any time we "try" something - a new recipe, a different tennis racquet, a new pair of shoes - we are testing. We are formulating a judgment about something on the basis of a sample of behavior. Language learners are testing their newly acquired forms of language almost every time they speak. They devise hypotheses about how the language forms are structured and how certain functions are expressed in forms. On the basis of the feedback they receive, they make judgments and decisions. Language teachers also test, informally and intuitively, in every contact with learners. As a learner speaks or writes or indicates either aural or reading comprehension, the teacher makes a judgment about the performance and from that judgment infers certain competence on the part of the learner. Classroom-oriented informal testing is an everyday and very common activity in which teachers engage almost intuitively.
A test, in plain, ordinary words, is a method of measuring a person´s ability or knowledge in a given area. The definition captures the essential components of a test. A test is first a method. There is a set of techniques, procedures, test items, that constitute an instrument of some sort. And that method generally requires some performance or activity on the part of either the testee or the tester or both. The method may be quite intuitive and informal, as in the case of judging offl1and someone's authenticity of pronunciation. Or it may be quite explicit and structured, as in a multiple-choice technique in which correct responses have already been specified by some "objective" means.

Next, a test has the purpose of measuring. Some measurements are rather broad and inexact while others are quantified in mathematically precise terms. The difference between formal and informal testing exists to a great degree in the nature of the quantification of data. Informal tests, the everyday intuitive judging that we do as laypersons or teachers, are difficult to quantify. Judgments are rendered in rather global terms. For example, it is common to speak of a "good" tennis player, "fair" performance by an actor in a play, or a "poor" reader. In formal testing, in which carefully planned techniques of assessment are used, quantification is important, especially for comparison either within an individual (say, at the beginning and the end of a course) or across individuals.
A test measures a person's ability or knowledge. Care must be taken in any test to understand who the testees are. What is their previous experience- their entry behavior? Is the test appropriate for them? How are scores to be interpreted for individuals? '

Also being measured in a test is, ability or knowledge-that is, competence. A test samples performance but infers certain competence. A driving test for a driver's license is a test requiring a sample of performance, but that performance is us ed by the tester to infer someone's general competence to drive a car. A language test samples language behavior and infers general ability in a language. A test of reading comprehension may consist of some questions following one or two paragraphs, a tiny sample of a second language learner's total reading behavior. From the results of that test the examiner infers a certain level of general reading ability.
Finally, a test measures a given area. In the case of a proficiency test, even though the actual 'performance' on the test involves only a sampling of skills, that area is overall proficiency in a language-general competence, in all skills of a language. Other tests may have more specific criteria. A test of pronunciation might well be a test only of a particular phonemic minimal pair in a language. One of the biggest obstacles to overcome in constructing adequate tests is to measure the criterion and not inadvertently something else.

Practicality
The foregoing discussion implicitly dealt with the three requirements of a "good" test: practicality, reliability, and validity. If these three axiomatic criteria are carefully met, a test should then be administerable within given constraints, be dependable, and actually measure what it intends to measure.
A test ought to be practical-within the means of financial limitations, time constraints, ease of administration, and scoring and interpretation. A test that is prohibitively expensive is impractical. A test of language proficiency that takes a student 10 hours to complete is also impractical. A test that requires individual one-to-one proctoring is impractical for a group of 500 people and only a handful of examiners. A test that takes a few minutes for a student to take and several hours for the examiner to correct is impractical for a large number of testees and one examiner if results are expected within a short time. A test that can be scored only by computer is impractical if the test takes place a thousand miles away from the nearest computer. The value and quality of a test are dependent upon such nitty-gritty, practical considerations.