- Home
- e-Journals
- Toegepaste Taalwetenschap in Artikelen
- Previous Issues
- Volume 31, Issue, 1988
Toegepaste Taalwetenschap in Artikelen - Volume 31, Issue 1, 1988
Volume 31, Issue 1, 1988
-
Taalvaardigheid: Meten Versus Observatie En Analyse
Author(s): T. v.d. Geestpp.: 25–33 (9)More LessDifferent types of validation are subjected to a critical examination, notably the underlying inadequate psychological latent trait' personality-theory and the unacceptably low prognostic capacity (or external or criterion related validity) of tests.It is argued that by the application of the Rasch-Analysis the results of tests are even worse in terms of explanatory adequacy with respect to content and criterion related validity, although esthetically the quantitative results look better.In the second part of the paper three different investigations are discussed. It is demonstrated that test results of these studies do not correlate or even correlate negatively with the computed results of an observational study of foreign (or second) language capacity. It is concluded that tests should not be used in order to determine proficiency.
-
Toetsing In Taalverwervingsonderzoek: Een Ondergeschoven Kindje?
Author(s): P.J.M. Grootpp.: 34–44 (11)More LessWhereas in the exact sciences concern for the measuring instruments used to gather research data is common practice, negligence of this aspect frequently occurs in the softer disciplines like language acquisition studies. This article deals with some examples of studies where faulty conclusions were drawn because the research data were unreliable due to the above mentioned neglect of the instruments used.
-
Bootstraps, Een Nieuw Statistisch Alternatief
Author(s): E.D.J. Schilspp.: 45–56 (12)More LessThe bootstrap is a Monte Carlo method for the approximation of the sampling error of a statistic. The bootstrap method estimates this standard error on the basis of the repeated calculation of the statistic at hand in each of a great number of so-called bootstrap samples, i.e. samples with replacement from a probability distribution which exactly mirrors the empirical relative frequency distribution. The method is useful when there is no analytical sampling theory available for the statistic at hand, or when violation of underlying assumptions precludes the application of an available sampling theory.This paper uses an analytically well-known problem as the context for the presentation of the method, viz. the sampling distribution of the arithmetic mean. The method is then applied to an investigation of language loss using an unfamiliar research design.
-
Le Modele De Rasch: Les Principes Sous-Jacents Et Son Application a La Validation De Tests
Author(s): John H.A.L. de Jongpp.: 57–70 (14)More LessThis paper provides an elementary introduction to the one parameter psychometric model known as the Rasch model. It explains the basic principles underlying the model and the concepts of unidimensionality, local stochastic independence, and additivity in non-mathematical terms. The requirements of measurement procedures, the measurement of latent traits, the control on model fit, and the definition of a trait are discussed. It is argued that the Rasch model is particularly appropriate to understand the mutual dependence of test reliability and validity. Examples from foreign language listening comprehension tests are used to illustrate the application of the model to a test validation procedure.
-
Padanalyse En Causale Modellering
Author(s): Leo J.Th. van der Kamppp.: 71–80 (10)More LessPad analysis is part of multivariate analysis and aims at describing the relation between various measurements in terms of a causal model which can be derived from a theory about the possible relations between these variables. Such a causal model can be presented in the form of an arrow diagram or pad diagram. On the basis of three examples the ingredients of a pad model are discussed, as well as the aim of pad analysis, its various forms (recursive vs. non-recursive models; models with manifest and latent variables; measurement errors). In addition causal analysis is discussed, a form of analysis which may be regarded as a generalization of pad analysis. Finally some comments are made on the state of the art and annotated references to the literature are given.
-
De Constructie Van Spreektaken In Grootschalig Peilingsonderzoek: Domeinbeschrijving En -Verkaveling
Author(s): Gert Rijlaarsdampp.: 81–90 (10)More LessIn this article a survey is given of the domain of oracy skills that will be the subject of a periodical repeated nationwide assessment study.The domain for the assessment of eleven-year-old pupils is restricted by two principles.a. The assessment includes communicative behaviour in testing situations which approach every day situations (school as well as non-school) that pupils of that age might find themselves in.b. The description of the skills will be given in so-called 'curricular significant units', units that are recognized by teachers as teachable units.The internal structure of the domain is dominated by two dimensions: the dimension of language functions and the dimension of skills. The dimension of language functions is restricted to transactional language. Distinctions are made between informative (reporting and expository) and conative (regulative and argumentative) subfunctions. These four subfunctions are used as classification units for speech acts. The speech functions and acts are the gists of language situations, generated and formulated by assessors and relevant respondents from various fields. In the dimension of oracy skills, five skills are distinguished:a. to select the information from the available information in the tasks and context;b. to arrange the information in an audible manner;c. to interact with listeners or other participants on the level of content (providing a context for instance) and relationship (termtaking);d. to formulate correctly and fluently;e. to use channel specific features approppriately: volume, tempo, pronunciation, etc.When these two dimensions are crossed, the matrix of curricular significant units is created. For each of the four language functions from three up to six tasks are constructed. Rating will be done via analytical schemes for each skill and by wholistic judgement by trained raters.
-
Het Gebruik Van Gestructureerde En Correctieopdrachten In Peilingsonderzoek Naar Schrijfvaardigheid
Author(s): Rob Schoonenpp.: 91–102 (12)More LessThe assessment of writing ability is problematic, insofar as it concerns reliable, valid and practical measurement. In this article the possibilities of reducing error in both the writer's production and the judge's scoring are discussed. Special attention is paid to the reduction of the writing process (as described by Hayes and Flower) by structuring the assignment. Reducing the writing process by eliminating subprocesses from the assignment results in a reduction of the number of possible sources of error in the writer's production. However, the validity of the structured writing assignments might be questioned.An example of a structured assignment and a revision task are given, and some preliminary results are discussed.
-
Tweede-Taalonderwijs Op Maat: Diagnostiserend Onderwijs Aan Volwassenen
Author(s): L. van den Boschpp.: 103–115 (13)More LessThe Institute for Educational Measurement (Cito) and Tilburg University have developed an instructional device for second language teaching to adults. It consists of:- a diagnostic language test- a set of teaching materialsThe diagnostic language test is based on a psycholinguistic model for language proficiency in which several linguistic levels are distinguished (Levelt & Kempen, 1976): phonological level, lexical level, morphological level, syntactic level and text level. The diagnostic language test consists of eleven subtests with which lexical, morphological, syntactic and textual skills can be examined. If test results show lack of proficiency, the additional set of teaching materials can be used in various ways to set up educational activities.
-
Kwaliteit En Efficientie Van Multiple-Choice Cloze-Toetsing
Author(s): W. Jochems and F. Montenspp.: 116–126 (11)More LessThis article presents and discusses a number of empirical findings concerning the psychometric quality of multiple-choice cloze tests as tests of general language proficiency, with emphasis on their validity and efficiency.The Dutch proficiency of various groups of foreign speakers was measured both by a series of separate proficiency tests in listening, speaking, reading and writing and by a series of multiple-choice cloze tests. Scores on multiple-choice cloze tests were found to correlate significantly with those on each of the proficiency tests. In addition, scores on multiple-choice cloze tests appeared to form a solid basis for predicting the total scores for listening, speaking, reading and writing taken together. Further, a clear relation was found to exist between levels of language proficiency and subjects' scores on multiple-choice cloze tests. Our conclusion is that the multiple-choice cloze tests under investigation have proved to be high-quality instruments for measuring proficiency in Dutch as a second language.Compared to a four-skills test, a multiple-choice cloze test is a very efficient instrument. Administering and processing take only little time. Besides, they can be administered to very large groups of subjects. Because of its quality and efficiency, multiple-choice cloze testing should be preferred to four-skills testing.
-
Kunnen We Volstaan Met Lakmoes?
Author(s): J.M. Janssen-van Dietenpp.: 127–137 (11)More LessIn order to be admitted to a Dutch university, foreign students have to take a Dutch proficiency test. It is argued by one of the Dutch universities that this proficiency can be demonstrated by a single type of test: a multiple-choice cloze test. This position is accounted for by concurrent validation and is justified by the efficiency criterion. In the present article this concurrent validity is questioned and it is argued that other considerations should play a role as well in evaluating tests, the results of which have far reaching consequences for individuals.
-
Eurocert: An International Certificate For english
Author(s): Jan van Weerenpp.: 142–148 (7)More LessThe EUROCERT English Proficiency Certification Program is jomtly developed and sponsored by the Educational Testing Service (ETS) and the Dutch Institute for Educational Measurement (Cito), and will be administered throughout Europe by Cito. It measures English proficiency - specifically listening, reading, writing and speaking, and utilizes the Test of English as a Foreign Language (TOEFL), the Test of Written English (TWE) and the Test of Spoken English (TSE). There are two levels of certification: Certified and Certified with Honors.EUROCERT measures proficiency at an advanced level. The different components of EUROCERT, i.e. the three-part TOEFL, the compound TSE and the essay-type TWE are discussed. Evaluation of language proficiency seems to be possible firstly by observing authentic language behaviour and secondly by an examination of the tacit knowledge that underlies language behaviour. It is not possible to make a choice between the two approaches on mere theoretical grounds. Generally, it is recommended to incorporate different techniques of measurement in a test in order to avoid method effects. The EUROCERT battery tries to meet this requirement.
-
Spreekvaardigheidstoetsing: Een Vergelijking Van Discussie-Opdrachten En Geleide Test
Author(s): L.J.A. Nienhuispp.: 149–159 (11)More LessAt higher levels of second language proficiency, the speaking skill is frequently measured in interviews, but the use of guided tests and group discussions is common too and can improve efficiency. Guided test and discussion create rather different speech situations and might well elicitate dissimilar kinds of oral production: the informal setting of group discussion is certainly less 'anxiety-provoking' and elicitates more natural speech, but it may lead to the use of elementary and unmonitored, minimally correct speech.In this article we report on a small-scale empirical investigation intended to lay bare differences between the language used in a guided test and in a group discussion at the level of 1st year university students of French.Although the discussion subject was defined in such a way that it would enable students to discuss a rather wide range of aspects, even superficial analysis of guided test and discussion subject suggested the first to be more content valid: this quality manifested itself in the higher proportion of different words used in the guided test answers. In other respects, there were no systematic differences between test and discussion: proportions of unique words and of less frequent words, in relation to the number of different words, were nearly the same in the two kinds of speech production. Contrary to intuitive expectation, the number of lexical and grammatical errors was greater in test production (i.e. in the formal setting) than in discussion, the guided test being perhaps a more demanding task, but the possible conclusion that the overall quality of second language use was less good in the guided test was not supported by other findings: mean scores of the three raters did not show systematic differences between test and discussion.Correlations between test scores and discussion scores were about .77, suggesting that as tests of speaking proficiency when the criterion is correctness, guided test and discussion are not as different as they may seem.The main difference between the two is in rater reliability: interrater correlations for the test were about .82; for the discussion the mean of three correlations was .52, but two of them approximated .60. One of the problems of rating discussions may be the rather unequal participation of the members of the group.The quality of discussion as a speaking proficiency test can, in our opinion, be improved by defining its subject in such a way that the aspects discussed will be sufficiently diverse and by training or instructing students: they should all participate actively and pay attention to regular turn-taking.
-
Tekstbegrip In...?
Author(s): Jan Metspp.: 160–171 (12)More LessFor a number of years Dutch foreign language examinations have comprised tests of reading comprehension in only one format: the m.c. question with four options. Now research is being carried out into the feasibility of adding another format: a modified cloze in which deletions are not random but deliberate and in which options are offered. Through these modifications a number of problems adhering to the regular cloze are avoided, and at the same time it is possible to maintain the concept of reading comprehension which has thus far governed the construction of FL examinations, the idea, that is, that reading comprehension is a unitary skill to be measured by questions requiring testees to integrate related information derived from text units larger than the sentence and often from diverse elements in the text.As it was precisely this that experts doubted cloze could measure, the central question to be answered was: can an 'm.c./ fill-in tesť be constructed which appeals to the same skill as the traditional m.c. format, but which may offer certain advantages (such as greater efficiency).First of all, the new test format was applied to two texts taken from a regular exam of reading comprehension in EFL, resulting in two tests of 26 items each - in the exam the texts had been accompanied by 11 and 12 questions respectively. Next, in a linguistic analysis, the relations between the items and the text were specified. It was found that each item in the two tests required more understanding of the text than just of the immediate environment of the deletion. The specifications were then submitted to the judgement of language testing experts at Cito, who confirmed the findings in the vast majority of cases. Subsequently, the new tests and the original ones were administered to a population of over 500 Dutch students in order to tests these hypotheses:1. The psychometric qualities of the m.c./fill in tests are just as good as those of the conventional tests, if not better.2. The items in the m.c/fill in tests appeal to the same skill as the conventional m.citems.3. The m.c./fill in tests are more difficult than the corresponding conventional tests.Hypotheses 1 and 3 were confirmed, allowing these tentative conclusions: the new test format is more efficient than the traditional one and offers testees less undue support in understanding texts. The second hypothesis was not confirmed, as the Rasch model for the three combinations of tests had to be rejected on formal grounds. However, as the model did not constitute an altogether inadequate description of the data, and as combining the various test formats did not affect the measure of fit negatively, it is still hoped that further analysis may show that the two formats do not appeal to fundamentally different dimensions.
-
Toetsing Tekstbegrip Moderne Vreemde Talen Door Middel Van Een Samenvattingsopdracht
Author(s): Leijn Melsepp.: 172–181 (10)More LessCan reading comprehension be measured in a valid way by having pupils, at the higher levels of secondary education, write a précis? An attempt to answer this question was made with respect to German and French texts. An advantage of setting a précis is that it does without comprehension questions. Such questions have the undesirable effect of facilitating the interpretation of a text. Another argument in favour of précis writing is its practical relevance (e.g. for later studies). A disadvantage is the inevitably subjective element in the rating. If this subjectivity cannot be kept within certain bounds, précis writing does not qualify for inclusion in a centrally organized examination. One aim of the present research was to establish the extent of interrater-reliability in the assessment of précis. Preliminary to this research it had to be determined what sort of précis would be most suitable. Its main features were defined as follows: its length had to be about a quarter of the original text, it had to be an objective rendering of the text, but pupils would not be required to adhere to the order of its parts. Rating instructions were produced giving the criteria for assessment and specifying the assignment of score marks. Research was done to establish the measure of agreement between ratings in which these instructions were observed and between independent assessments. The ratings on the basis of the instructions showed a higher measure of agreement than independent assessments. This research was repeated after the instructions had been revised leading to a substantial improvement of agreement between ratings.
-
Communicatieve Opgaven In De Examens Tekstbegrip Moderne Vreemde Talen
Author(s): I.F. Wijghpp.: 182–186 (5)More LessSince 1987 the tests of reading comprehension at the lower vocational and secondary school level include a dozen communicative assignments. The languages concerned are French, German, English and Spanish. These assignments have the following characteristics:- the texts differ in length: they are shorter than the traditional texts;- the texts are printed in their original lay-out;- their is a large variety in types of text;- the texts are 'consumer-texts': foreigners will probably come into contact with such texts during their stay in the country where the object language is spoken;- there is usually only one question per text;- the questions are meant to call up language performance that goes with a given type of text in reality; i.e. they tie in with the reading objective that is most functional;Cito (the Dutch national institute for educational measurement) has done research into the possibilities of including communicative assignments in the exams at the higher levels of secondary education. Basic principles have been laid down and a comparison with the assignments at lower levels has been made. Experimental tests have been constructed; they have been taken by pupils at the levels concerned. The psychometric data have not been such that a decision on implementation could be made. Further research and decisions will be necessary.
Volumes & issues
-
Volume 86 (2011)
-
Volume 84 (2010)
-
Volume 83 (2010)
-
Volume 84-85 (2010)
-
Volume 82 (2009)
-
Volume 81 (2009)
-
Volume 80 (2008)
-
Volume 79 (2008)
-
Volume 78 (2007)
-
Volume 77 (2007)
-
Volume 76 (2006)
-
Volume 75 (2006)
-
Volume 74 (2005)
-
Volume 73 (2005)
-
Volume 72 (2004)
-
Volume 71 (2004)
-
Volume 70 (2003)
-
Volume 69 (2003)
-
Volume 68 (2002)
-
Volume 67 (2002)
-
Volume 66 (2001)
-
Volume 65 (2001)
-
Volume 64 (2000)
-
Volume 63 (2000)
-
Volume 62 (1999)
-
Volume 61 (1999)
-
Volume 60 (1998)
-
Volume 59 (1998)
-
Volume 58 (1998)
-
Volume 57 (1997)
-
Volume 56 (1997)
-
Volume 55 (1996)
-
Volume 54 (1996)
-
Volume 53 (1995)
-
Volume 52 (1995)
-
Volume 51 (1995)
-
Volume 50 (1994)
-
Volume 49 (1994)
-
Volume 48 (1994)
-
Volume 45 (1993)
-
Volume 46-47 (1993)
-
Volume 44 (1992)
-
Volume 43 (1992)
-
Volume 42 (1992)
-
Volume 41 (1991)
-
Volume 40 (1991)
-
Volume 39 (1991)
-
Volume 38 (1990)
-
Volume 37 (1990)
-
Volume 36 (1990)
-
Volume 35 (1989)
-
Volume 34 (1989)
-
Volume 33 (1989)
-
Volume 32 (1988)
-
Volume 31 (1988)
-
Volume 30 (1988)
-
Volume 29 (1987)
-
Volume 28 (1987)
-
Volume 27 (1987)
-
Volume 26 (1986)
-
Volume 25 (1986)
-
Volume 24 (1986)
-
Volume 23 (1985)
-
Volume 22 (1985)
-
Volume 21 (1985)
-
Volume 20 (1984)
-
Volume 19 (1984)
-
Volume 18 (1984)
-
Volume 17 (1983)
-
Volume 16 (1983)
-
Volume 15 (1983)
-
Volume 14 (1982)
-
Volume 13 (1982)
-
Volume 12 (1982)
-
Volume 11 (1981)
-
Volume 10 (1981)
-
Volume 9 (1981)
-
Volume 8 (1980)
-
Volume 7 (1979)
-
Volume 6 (1979)
-
Volume 5 (1978)
-
Volume 4 (1978)
-
Volume 3 (1977)
-
Volume 2 (1977)
-
Volume 1 (1976)
Most Read This Month
Article
content/journals/22134883
Journal
10
5
false
