The last type that could be discussed is criterion-related validity. Weir (1990:22.) assumes that it is connected with test scores link between two different performances of the same test: either older established test or future criterion performance. The author of the paper considers that this type of validity is closely connected with criterion and evaluation the teacher uses to assess the test. It could mean that the teacher has to work out definite evaluation system and, moreover, should explain what she finds important and worth evaluating and why. Usually the teachers design their own system; often these are points that the students can obtain fulfilling a certain task. Later the points are gathered and counted for the mark to be put. Furthermore, the teacher can have a special table with points and relevant marks. According to our knowledge, the language teachers decide on the criteria together during a special meeting devoted to that topic, and later they keep to it for the whole study year. Moreover, the teachers are supposed to make his/her students acquainted with their evaluation system for the students to be aware what they are expected to do.
According to Bynom (Forum, 2001) reliability shows that the tests results will be similar and will not change if one and the same test will be given on various days. The author of the paper is of the same mind with Bynom and presumes the reliability to be the one of the key elements of a good test in general. For, as it has been already discussed before, the essence of reliability is that when the students scores for one and the same test, though given at different periods of time and with a rather extended interval, will be approximately the same. It will not only display the idea that the test is well organized, but will denote that the students have acquired the new material well.
A reliable test, according to Bynom, will contain well-formulated tasks and not indefinite questions; the student will know what exactly should be done. The test will always present ready examples at the beginning of each task to clarify what should be done. The students will not be frustrated and will know exactly what they are asked to perform. However, judging form the personal experience, the author of the paper has to admit, that even such hints may confuse the students; they may fail to understand the requirements and, consequently, fail to complete the task correctly. This could be explained by the fact that the students are very often inattentive, lack patience and try to accomplish the test quickly without bothering to double check it.
Further, regarding to Heaton (1990:13), who states that the test could be unreliable if the two different markers mark it, we can add that this factor should be accepted, as well. For example, one representative of marking team could be rather lenient and have different demands and requirements, but the other one could appear to be too strict and would pay attention to any detail. Thus, we can come to another important factor influencing the reliability that is markers comparison of examinees answers. Moreover, we have to admit a rather sad fact but not the exceptional one that the makers personal attitude towards the testee could impact his/her evaluation. No one has to exclude various home or health problems the marker can encounter at that moment, as well.
To summarize, we can say that for a good test possessing validity and reliability is not enough. The test should be practical, or in other words, efficient. It should be easily understood by the examinee, ease scored and administered, and, certainly, rather cheap. It should not last for eternity, for both examiner and examinee could become tired during five hours non-stop testing process. Moreover, testing the students the teachers should be aware of the fact that together with checking their knowledge the test can influence the students negatively. Therefore, the teachers ought to design such a test that it could encourage the students, but not to make them reassure in their own abilities. The test should be a friend, not an enemy. Thus, the issue of validity and reliability is very essential in creating a good test. The test should measure what it is supposed to measure, but not the knowledge beyond the students abilities. Moreover, the test will be a true indicator whether the learning process and the teachers work is effective.
Types of tests
Different scholars (Alderson, 1996; Heaton, 1990; Underhill, 1991) in their researches ask the similar question why test, do the teachers really need them and for what purpose. Further, they all agree that test is not the teachers desire to catch the students unprepared with what they are not acquainted; it is also not the motivating factor for the students to study. In fact, the test is a request for information and possibility to learn what the teachers did not know about their students before. We can add here that the test is important for the students, too, though they are unaware of that. The test is supposed to display not only the students weak points, but also their strong sides. It could act as an indicator of progress the student is gradually making learning the language. Moreover, we can cite the idea of Hughes (1989:5) who emphasises that we can check the progress, general or specific knowledge of the students, etc. This claim will directly lead us to the statement that for each of these purposes there is a special type of testing. According to some scholars (Thompson, 2001; Hughes, 1989; Alderson, 1996; Heaton, 1990; Underhill, 1991), there are four traditional categories or types of tests: proficiency tests, achievement tests, diagnostic tests, and placement tests. The author of the paper, once being a teacher, can claim that she is acquainted with three of them and has frequently used them in her teaching practice.
In the following sub-chapters we are determined to discuss different types of tests and if possible to apply our own experience in using them.
3.1. Diagnostic tests
It is wise to start our discussion with that type of testing, for it is typically the first step each teacher, even non-language teacher, takes at the beginning of a new school year. In the establishment the author of the paper was working it was one of the main rules to start a new study year giving the students a diagnostic test. Every year the administration of the school had stemmed a special plan where every teacher was supposed to write when and how they were going to test their students. Moreover, the teachers were supposed to analyse the diagnostic tests, complete special documents and provide diagrams with the results of each class or group if a class was divided. Then, at the end of the study year the teachers were demanded to compare the results of them with the final, achievement test (see in Appendix 1). The author of the paper has used this type of test for several times, but had never gone deep into details how it is constructed, why and what for. Therefore, the facts listed below were of great value for her.
Referring to Longman Dictionary of LTAL (106) diagnostic tests is a test that is meant to display what the student knows and what s/he does not know. The dictionary gives an example of testing the learners pronunciation of English sounds. Moreover, the test can check the students knowledge before starting a particular course. Hughes (1989:6) adds that diagnostic tests are supposed to spot the students weak and strong points. Heaton (1990:13) compares such type of test with a diagnosis of a patient, and the teacher with a doctor who states the diagnosis. Underhill (1991:14.) adds that a diagnostic test provides the student with a variety of language elements, which will help the teacher to determine what the student knows or does not know. We believe that the teacher will intentionally include the material that either is presumed to be taught by a syllabus or could be a starting point for a course without the knowledge of which the further work is not possible. Thus, we fully agree with the Heatons comparison where he contrasts the test with a patients diagnosis. The diagnostic test displays the teacher a situation of the students current knowledge. This is very essential especially when the students return from their summer holidays (that produces a rather substantial gap in their knowledge) or if the students start a new course and the teacher is completely unfamiliar with the level of the group. Hence, the teacher has to consider carefully about the items s/he is interested in to teach. This consideration reflects Heatons proposal (ibid.), which stipulates that the teachers should be systematic to design the tasks that are supposed to illustrate the students abilities, and they should know what exactly they are testing. Moreover, Underhill (ibid.) points out that apart from the above-mentioned the most essential element of the diagnostic test is that the students should not feel depressed when the test is completed. Therefore, very often the teachers do not put any marks for the diagnostic test and sometimes even do not show the test to the learners if the students do not ask the teacher to return it. Nevertheless, regarding our own e