What do researchers/authors say about what SHOULD be tested at Uni in these subjects/what does the research or theorists say about what content should be taught at Uni and how…? All this you will later compare and contrast with your own data in the Discussion and Conclusion section of your TfG, so it is important to get as many contrastable opinions as possible now.
The purpose of this TFG is to see how English language should be tested at university level. It aims to figure out what type of test would benefit the most to Ingl?s Instrumental in Spanish universities, taking into account the reliability and validity of itself.
1 WHAT IS A TEST?
To begin with, it is essential to understand what a test is, according to the Cambridge Dictionary (2019), a test is a way of discovering, by questions or practical activities, what someone knows, or what someone or something can do or is like.
Having provided a general definition of what a test is, now it is time to present the different kinds of tests that could be related to Ingl?s Instrumental. Many studies have been carried out in relation to what kind of testing is more reliable and valid in the last decades, taking into account the kind of test and the form of assessing the test. However, it is also true that there is a process to follow that teachers can use in order to test students knowledge before, during and after the course.
1.1 Testing the previous knowledge
According to Soler (2008), testing the previous knowledge is instrumental for the assessment of specific objectives, as it allows teachers to determine the baseline reference point of the students, which will be later compared with their final status. The author asserts that it can be applied at the beginning of each unit. The attitude shown by this initial test is significantly important to determine the teaching approach during the course. Some books may have questionnaires that aim to identify these attitudes but teachers can also create their own questionnaires asking about relevant aspects that they may consider useful for the development of the course, related to students previous experiences with the English language. Soler states that previous knowledge should also be tested in different ways. In books, it can be found some grammar revision which was supposed to be taught before. Nevertheless, listening, writing, reading and speaking abilities are not usually included in these books revisions. Therefore, teachers should test them, and this can be done through written and oral-self introductory presentations that students will perform or listening to the radio, among other things.
1.2 Continuous assessment
Soler also states that the other way of evaluating related to the process is the continuous assessment. By doing this kind of testing, the assimilation of objectives can be tested regularly, and it gives a proof of the students progress. In the authors opinion, this type of assessment encourages continuous study, and it can be assessed by giving a note to their second test students realised. This is done for the students to see both process and final product. In the specific case of speaking, the author claims that the assimilation of objectives can be evaluated with students participation in class and application of new structures and relevant vocabulary to speaking activities.
2 TYPES OF TESTS
2.1 Final achievement tests
Once seen the two different processes of assessing, it is time to focus on what type of test can adjust more to the teachers needs. The first sorts of tests Hughes (2003) presents are the final achievement tests. These kinds of tests have two approaches directly related to language courses, and they are administered at the end of the course of study. They can be managed and written by ministries of education, official examining boards, and even by members of teaching institutions. The first approach, in the view of some testers, is the syllabus-content approach, that is to say, the content of a final achievement test, which is directly based on an accurate course syllabus. The main disadvantage of this approach is that if the syllabus is inadequately designed, it means that the materials chosen for this course were badly chosen and subsequently, the results could be misleading. An alternative approach related to this kind of test asserts that the aim of this test is to base its content straightforwardly on the objectives of the course.
This approach has several advantages, the first one is that course designers are pressed to be explicit about the objectives students should achieve and the second one is that this approach makes it possible for performance on the test and that will after show just how far students have achieved those objectives. Hughes also states that this type of tests work against the perpetuation of poor teaching practice, something that course-content based tests fail to do.
2.2 Proficiency tests
However, Hughes (2003) shows that not all the tests are related to language course, he explains that there are other sorts of tests which can be carried out without being based on it. This other variety of exams are based on other students qualities. As for instance, proficiency tests are a type of tests designed to measure students ability disregarding any previous training they might have had in that second language. In most of the cases, this type of exams is done for particular purposes, as its name says. Some examples of the use of these tests can be seen in a United Nations, for being a translator, or in a British university so that you can enter in a language course.
2.3 Diagnostic tests
Apart from the tests already mentioned, there are other kinds of tests, which despite not being directly related to Ingl?s Instrumental, they are important to mention for their relevance in testing. These tests Hughes refers to are diagnostic tests and placement tests. In the case of diagnostic tests, they are used to identify the students strengths and weaknesses, making sure of what further teaching will be more appropriate. At the level of wide language skills, this test is fairly straightforward but it is not that easy when the researcher wants to obtain a detailed analysis of the grammatical structures used by the learner. As for instance, if the teacher wants to know whether the student has acquired the present perfect. Another disadvantage of these kinds of tests is that a single example would not be enough to test the students, and therefore, a comprehensive diagnostic test would be vast and impractical for being a routine. Nevertheless, these tests can be significantly useful for individualised instruction or self-instruction where they can be better practiced and exemplified (Hughes, 2003).
2.4 Placements tests
Respecting placement tests, the author claims that they are intended to provide information, which will later place students at an appropriate level of their teaching programme regarding their abilities. Even though they can be bought, Hughes does not recommend doing it, unless the institution concerned is quite sure that the test considered suits with its particular teaching programme. Moreover, the idea of being commercially available implies that it will not work well for every interested institution.
As diagnostics tests, placement tests are made for particular situations, which depend on the identification of the key features of each level (Hughes, 2003).
3 TYPES OF TESTING
3.1 Direct versus indirect testing.
Once it has been seen all the kinds of tests Ingl?s Instrumental can be based on, it is the moment to pay particular attention to the types of testing. The classification of types of testing is going to help the teacher making him, or her know how to construct a test. The first type of testing Hughes (2003) mentions is direct versus indirect testing. On the one hand, testing is said to be direct when it requires the student to perform correctly what the instructor wants to measure. The tasks to be done should be as authentic as possible. This form of testing has a number of advantages. The first one is that it is clear about what abilities it is wanted to be assessed and therefore, it is relatively straightforward to create the conditions to do it. Moreover, in the case of productive skills, the assessment and interpretation are also easy in the authors view. And last but not least, it has a repercussion effect in the student, as hes practising his language skills at the same time that hes performing the test.
On the other hand, indirect testing seeks to measure the skills which underlie the abilities the examiner is keen on. One of the main advantages that this type of testing has is that it offers the possibility of testing a representative sample of a predetermined number of skills but underlying a potentially indefinitely large number of manifestations. By contrast, its relationship between performance on them and performance of the skills is rather feeble in strength and uncertain in nature. Although direct tests are usually easier to make, many testers are reluctant to dedicate themselves entirely to direct testing and will always include some indirect parts in their tests.
In order to better understand this form of testing, it will be exposed to the case of Shohamy (1992) in his form of seeing writing tests. Although both forms of testing can be applied to this skill, he explains throughout his study why he prefers direct testing for this specific ability. On the one hand, he asserts that indirect testing was used before in the writing proficiency indirectly, that is to say, the tasks required on these tests were not authentic. The main advantage that they had is that they involved objective elements to tests, such as multiple choice, and they could be easier scored. The reason for using indirect writing tests was due to the high correlation researchers found between indirect and direct tests. However, this correlation was not enough to assert that indirect tests measure the same features as those of direct tests. For example, an exercise of correcting mistakes of sentences, even if it is related to the process of writing, it is not a real writing exercise. This is the reason why direct tasks are considered valid since they reflect the actual task students might perform in real life. In the authors opinion, the main problem of direct tests is to design criteria to evaluate the direct writing samples done by the students.
Another fact to take into consideration is the reliability of direct testing. According to Shohamy (1992), there are three factors that affect it. The first one is the background of the rater (e.g. Teachers may be influenced by the instructional goals of the written sample they are given before, so they could be influenced by this, in the moment of assessing the test). The second factor is related to the training of the rater, in other words, teachers must receive intensive training in procedures for rating written proficiency. The third and final factor is based on the rating scale, which is the tool that provides the criteria for determining the quality of the written language sample.
3.2 Discrete point versus integrative testing.
Hughes (2003) also makes a distinction between discrete point and integrative testing. In his study, he refers to discrete point testing as the assessment of one element at a time, that is to say, item by item. This form of testing can be used to assess particular grammatical structures. Secondly, integrative testing, as opposed to discrete point testing, requires the student to use many language elements at the same time.
To have a better understanding of this sort of testing, it will be presented the results of a study done by Farhady (1979) related to this subject. Farhady argues in his study that there are a number of implications to choose between these two forms of testing. First of all, according to the authors opinion, if a test is only composed by either discrete or integrative points of testing, learners (regardless their proficiency level but determined by their previous educational knowledge) will perform it better or worse getting depending on an arbitrary result that will place him or her into an inappropriate level.
Another area where he believes that his research could be applied, it is language acquisition. Among all the different models of second language acquisition, he proposes the Monitor Model of Krashen (1997 a, 1997b) to explain the application of his research. In the Monitor Model theory, there was a distinction made between language acquisition (the unconscious process by which learners acquire language in both natural and classroom situations) and language learning (the conscious process that learners perform only through formal instruction). Krashen claimed that if learners were provided with enough time and focus on the form of the utterance, then, they would monitor their results of the performance. Therefore, for Farhady (1979), the Monitor Model predicts that discrete-point tests would be monitored by language learners, and integrative ones would not, supporting at the same time his own research.
Nevertheless, if there is no practical evidence between both sorts of testing, it would diminish one of the arguments based on prediction of the Monitor Model. And besides, this would imply a reconsideration of his own research considering this time, psychometric factors involved in both types of testing. However, this is why he recommends looking for such differences in language acquisition in other areas different from discrete point and integrative testing.
3.3 Norm-referenced versus criterion-referenced testing.
The third kind of testing Hughes (2003) mentions is nom-referenced and criterion-referenced testing. On the one side, norm-referenced testing relates ones student performance to that of other students. This type of testing is not interested in testing foreign language skills but instead, it places the candidate in the top ten per cent of the candidates who took the test, or on in the bottom five per cent, or it even says if the candidate performed better than the remaining sixty per cent of the class according his or her own score. By contrast, criterion-referenced testing directly provides information about what the student can actually do in language. They classify candidates simply by saying if they are able to perform the required task or not. By using this type of testing there are two befits: one is that set the standards of what can a learner actually do, and in addition to this, it can motivate the student to attain those standards. The only objection of this sort of testing is that its content does not let so much freedom to be changed, as norm-referenced tests do.
3.4 Objective versus subjective testing.
Hughes (2003) states that the distinction made by objective and subjective testing is related to the scoring and it is straightforward to follow, if no judgement is required from the scorer, then the scoring is objective if not, it is subjective. An example of objective testing could be a multiple choice test where correct answers are considered unambiguous. It is true that there are different degrees of subjectivity, and as it has been proved in the study of Shahomy (1992), where the author explained the reliability of writing tests. Another possibility Hughes mentions in order to be less subjective is to have greater agreement between two different scorers (that, in theory, will give two different approximate marks) or being the same scorer but correcting the same test in different occasions.
3.5 Communicative language testing.
The last type of testing that Hughes (2003) refers to is the communicative language testing. This type of testing simply calculates the ability to take parts in acts of communication, including reading and listening, and it explores the best way to do it. To conclude his research, he considers to have discussed the most predominant aspects in communicative testing and he rejects the idea of having a different heading for this recapitulation.
Dictionary.cambridge.org. (2019). Meaning of TEST in the Cambridge Dictionary. [online] Available at: [Accessed 2 Feb. 2019].
Farhady, H. (1979). The disjunctive Fallacy between discrete-point and integrative tests. TESOL Quarterly 13, 3, 355-357.
Hughes, A. (2003). Testing for language teachers. (2nd ed.). New York: Cambridge University Press, 9-21.
Krashen, S. (1977a). The monitor model for adult second language performance. In Burt, M. H. Dulay, and M. Finocchiaro ( Eds. ), Viewpoints on English as a second language. Regents Publishing Company, Inc. .
Krashen, S. (1977b). Some issues relating to the monitor model. In Brown, H. D. and C. Yorio ( Eds. ). On TESOL 77, TESOL, Washington.
Shohamy, E., Gordon, C. and Kraemer, R. (1992). The effect of raters background and training on the reliability of direct writing tests. Modern Language Journal 76, 2733.
Soler, LM (2008), Teaching English at University Level, Universitat Polit?cnica de Catalunya, Barcelona. Available from: ProQuest Ebook Central, 41-42.