Collocational Competence Of Iranian Learners Using Ctest English Language Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

One of the most significant current discussions in second or foreign language is collocations as a challenging attribute of second language learning and as a vital element of communicative competence. A crucial part of native speakers' communicative competence is collocational competence which can be defined as native speakers intuitively "know" which words usually come together and which do not. Moreover, any speech community has a set of idiomatic ways of stating ideas in certain complete phrases and a great many partly filled phrase-frames. From this it can be inferred that non-native speakers with deficient communicative competence have great difficulty in formulaic language. Therefore, the aim of this paper is to investigate the relationship between collocational competence and general language proficiency of Iranian students. To reach this goal, two tests were administered to Iranian learners, namely a cloze test and a c-test. The language proficiency level of the subjects was scrutinized through their performance on a cloze test. The subjects' collocational competence was calculated through their performance on a 50-item c-test consisting of verb noun collocations in which the verb missing but the first letter/phoneme was provided. The results of the statistical analyses demonstrate that there is a high relationship between collocational competence and general language proficiency of learners. It can be concluded that the c-test is an effective measurement to assess learners' collocational competence.

Keywords: Language proficiency, Collocation, Cloze test, C-test, Iranian EFL learners

Theoretical Background

The term "collocation" is discussed prevalently in many areas of linguistics, such as semantics, systematic linguistics, morpho-syntax, phraseology, corpus linguistics, and lexicography. Collocations are generally defined as words that "fit together" intuitively with great expectation in the syntagmatic and paradigmatic levels. Linguists studying collocations at the lexical level consider collocations as the linear and syntagmatic co-occurrence of lexical item (Firth, 1957; Halliday, 1966; Mcintosh, 1966; Mitchell, 1971; Sinclair, 1966). For example, putrid and rancid are synonymous, but putrid collocates with fish and rancid with butters. Collocations are also argued in terms of their syntactic restrictions (Benson, 1985; Greenbaum, 1970; Kijellmer, 1984; Mitchell, 1971; Nation, 2001), for instance, phrasal verbs like make up and semantic restrictions (Cowie & Mackin, 1975; Howarth, 1998; Lewis, 1997; Nation, 2001), as Good morning.

Collocational knowledge has been known as a vital element of native speakers' communicative competence (Moon 1992; Bahns and Eldaw 1993; Fontenelle 1994; Herbst 1996; Lennon 1996). Wray (2000, 2002a,b) argues that any speech community creates a set of idiomatic ways of stating their thought by favouring, merely through repeated use, some whole phrases and a great deal of partly filled phrase-frames. From this it can be assumed that non-native speakers with deficient communicative competence find formulaic language extremely difficult. Numerous investigators have maintained that collocational knowledge is a significant issue that contributes to the dissimilarities between native speakers and foreign language students (Aston, 1995; Fillmore, 1979; Kijellmer, 1991; Pawley & Syder, 1983). It is particularly when EFL learners fail to use collocations accurately; it is a chief sign of foreignness (McArthur, 1992; McCarthy, 1990; Nattinger, 1980; Wu, 1996). The most effective claim is that collocational competence is an essential part in the process of second/foreign language learning (Lewis, 1997, 2000; Nattinger & DeCarrico, 1992; Richards & Rogers, 2001). Cowie (1988), furthermore, claims that institutionalized units (lexical phrases and collocations) serve communicative needs and enable individuals to reuse and create the units. He indicates that stability and creativity of institutionalized units are complementary and interactive factors in vocabulary' use and suggests vocabulary teaching should keep a balance between lexical phrases and collocations. As Ellis (2001) argues that collocational knowledge is the essence of language competence.

Collocation Types

Researchers or lexicographers discriminate collocations into different types from various perspectives. The compilers of Oxford Dictionary of Current Idiomatic English, Cowie and Mackin (1975), use the idiomaticity to categorize collocations and idioms into four groups according to idiomaticity from most to least fixed: pure idioms, figurative idioms, restricted collocations, and open collocations. Pure idioms, such as blow the gaff, are the most fixed with semantic opacity. Figurative idioms, such as catch fire and a close shave, are more transparent and not quite fixed. Restricted collocations are collocations that have one element used in a non-literal sense and the other used in its normal meaning, like, under a shadow, while open collocations involve elements which are freely combinable and each element has its literal sense such as mad dog. Howarth (1998) also regards free combinations as collocations and categorises collocations into free collocations and restricted collocations according to their degree of restrictedness.

On the other hand, Benson et al. (1997) in The BBI Dictionary of English Word Combinations illustrate two types of collocations by syntactic nature. The collocations fall into two categories: grammatical collocations and lexical collocations. A grammatical collocation is a phrase which consists of a content word (a noun, a verb, or an adjective), plus a grammatical word such as a preposition or grammatical structure, e.g. a clause or an infinitive. Examples include account for, by accident, argument about and so on. Lexical collocations, on the other hand, exclude prepositions, infinitives or clauses and contain various combinations of nouns, adjectives, adverbs, and verbs, such as affect deeply, bees buzz, and so on. The BBI Dictionary of English Word Combinations contains eight main types of grammatical collocations and seven types of lexical collocations. Nevertheless, the authors do not include "free combinations" in their dictionary which are, in effect, limitless.

Furthermore, Lewis (1997) indicates that collocations may be strong, weak, frequent or infrequent. Lewis divides collocations into strong and weak ones according to their fixedness and restrictedness, and distinguishes collocations from frequent and infrequent ones due to their frequency of co-occurrence in a corpus. Strong collocations are recognized as tightly linked phrases which almost function like single words. In contrast, weak collocations, such as a nice day and a good chance, are combinations of two common words, and each of which may occur with many other words. Collocations can also be any combination of strong and frequent, strong and infrequent, weak and frequent, or weak and infrequent. This study considered verb noun collocations which are the most frequent but difficult to master.


There are different tests to measure EFL language proficiency or communicative language usage. In Bachman's view (1990: 166), "the primary interest in using language tests is to make inferences about one or more components of an individual's communicative language ability".

The C-test has been proposed by Klein-Braley and Raatz (Klein-Braley & Raatz,1984; Klein-Braley, 1935). The beneficial aspects c-test is an easy to administer test format, fast and objective scoring, and high reliability. They are similar to the classic cloze test and are based on the principle of reduced redundancy testing (Klein-Braley, 1994; 1997). The numerous empirical evidences back a positive overall evaluation of C-tests as tests of general language proficiency.

For instance, Dörnyei and Katona (1992: 203) in their work on an English C-test stated:

The C-test proved to be a highly integrative and versatile measuring instrument, working well in samples of various difficulty and homogeneity levels . . . our conclusion about the C-test is that not only is a reliable and valid measure of general language proficiency, but it is also one of the most efficient language testing instruments in terms of the ratio between resources invested and measurement accuracy obtained.

Therefore, C-tests require the integration of both skills and knowledge: a core competence in all types of language use. Furthermore, the lexical competence needed in the processing of cloze tasks (including c-tests) as both knowledge of individual words and ability to use contextual clues to find out which word fits a blank (Read, 2000: 113). He discussed that 'a cloze tends to make a very embedded assessment of vocabulary; to the extent that it is difficult to unearth the distinctive contribution that vocabulary makes to test performance' (p. 115; stress in original).

In a traditional C-test, the second half of every key word in a phrase is deleted, leaving the first and last sentence of the passage intact. With the C-test, a clue (half the word) serves as a stimulus for respondents to find the other half. In this version of the C-test, only the first letter/phoneme of every deleted word was provided, for instance, "m_____ a mistake," " s______ coffee," or " ch _____ his mind" in order to elicit the specific collocations intended for this research and to lessen the chance of guessing by subjects. In this study, the C-test which was a modified version of Al-Zahrani (1998) used for collecting data.

The test items included only lexical collocations. A distinction has been made between lexical collocations and grammatical collocations (Benson & Benson, 1986, 1997). Lexical collocations do not have a dominant word; they have structures such as the following: verb + noun, adjective + noun, noun + verb, noun + noun, adverb + adjective, adverb + verb" (p: ix). In other words, a lexical collocation is more lexis-oriented but a grammatical collocation is more grammar-oriented. In this study, lexical collocations - verb noun- were measured.

The present study

A review of literature shows that there is not enough empirical research on the relationship between EFL learners' collocational knowledge and their performance on a C-test. However, a number of studies have discussed the relationships between language proficiency and collocational knowledge which is closely related to the topic of this study. Al-Zahrani (1988) found that there is a high correlation between language proficiency and collocational knowledge among EFL Saudi's students. Sung (2003) found that there was a significantly strong correlation between the EFL students' knowledge of lexical collocations and their speaking fluency.

The present study adds to and expands those previous studies; however, no study, to the best of my knowledge, has used a cloze test and a C-test to measure the relationship between language proficiency and collocational competence. Hence, the current study tries to fill this gap. It presents the following research question:

What is the relationship between students' language proficiency and the knowledge of verb noun collocations?



Participants in this study were English major students (44 male and 168 female) aged 18-35 in the English Department at Islamic Azad University, Tehran North Branch (IZUTNB) in Tehran. They participated in a cloze test, and a blank-filling collocation test. The cloze test was piloted with 30 subjects, but c-test was piloted twice: the first one with the same subjects who participated in the cloze test, the second one with different participants to reduce backwash effects.

Instruments and procedure

The instruments of this study were a cloze test and a c-test to measure the language proficiency and collolcational competence of the subjects. Cloze tests have been used to measure overall ability in a language (Hughes 1989, 62ff.). The procedure requires the test taker to fill in the blanks in a text in which a number of words have been deleted, usually at specific intervals. The deletion of words at regular intervals apparently produces a representative sample of the linguistic features of the text, which makes it possible to obtain a valid measure of the test-takers' underlying language ability.

The cloze test was developed by Suter (2002) was piloted on 30 students in IZUTNB. They were all English major students. The responses attained from the pilot study were cautiously reviewed. It was noted that some items were not in a suitable degree of difficulties for students, so they were modified. The papers were marked by semantically accepted word or SEMAC technique (see Suter, 2002). The reliability for the cloze test was estimated by Cronbach's alpha which was .78. The result shows that the cloze test has a high reliability and can be used to measure students' knowledge of English.

The c-test was developed by Al-Zahrani (1998) was piloted on the same 30 students in IZUTNB. They were all English major students. The responses attained from the pilot study were carefully reviewed. It was found out that some items were not in a suitable degree of difficulties for students; therefore, they were changed and piloted again on the 30 different students in IZUTNB. The reliability for the c- test was estimated by Cronbach's alpha which was .87. The result shows that the c-test has a high reliability and can be used to measure students' collolcational competence.

To validate the two modified tests of the study, they were administered to a different group of 30 students. The item facility and item discrimination were calculated. Items facility for the both cloze test and the test of collocations fell within the range of (.47 and .70 respectively) (cf. Farhady, 1995).

After the scatter plot had shown a linear relationship between students' collocational competence and their overall proficiency, the researcher used Pearson correlation to quantify the strength of such a relationship. Table 1shows the results of the correlations.

Table : Pearson Correlation of the Test of Collocation and the Cloze test

Test of Collocations

Cloze test .532**

**. Correlation is significant at the 0.01 level (2-tailed).

As can be seen in Table 1, the relationship between language proficiency of students (as measured by the cloze test) and the collocational competence (as measured by the c-test) was explored using Pearson product-moment correlation coefficient. Afterwards, coefficient of determination was calculated, and it is 28.324 and shows students' overall proficiency helps to explain 34 per cent of the variance in respondent scores on the collocational competence. Preliminary analyses were carried out to make sure no violation of the assumptions of normality, linearity and homoscedasticity. According to Cohen (1988) guideline r=.10 to .29 or r=-.10 to -.29 small, r=.30 to .49 or r=-.30 to -.4.9 medium, and r=.50 to 1.0 or r=-.50 to -1.0 large, there was a strong, positive correlation between the two variables [r=.532, n=212, p<.0005] with high levels of language proficiency associated with high levels of collocational competence. Therefore, the results show there is a high correlation between language proficiency and the collocational knowledge of students. My findings support the findings of Al-Zahrani (1988) and Hsu and Chui (2008).

Discussion and Pedagogical Implication

Regarding the research question, the results of the statistical analyses of the present study showed that there was a strong positive correlation between students' verb noun collocational competence as measured by the c-test and the language proficiency as measured by the cloze test. This finding is in line with the findings of Al-Zahrani (1988) and Hsu and Chui (2008). The results of Pearson correlation supported the claim that the development of collolcational knowledge was in line with other language skills. An implication of this finding is that although collocations are not taught to students explicitly at English and Translation Department in IZUTBN, they learn them implicitly. Another implication is that curricula used help students to develop their collolcational knowledge alongside of their overall language proficiency. It also supports this idea that it is important to teach language collocationally instead of individually (Farghal and Obidat, 1995; Howarth, 1998).

Bahns and Eldow (1993) propose that c-tests and cloze tests can be used to measure learners' language proficiency which is in line with the findings of this study. They also suggest that test designers should make and validate collocation tests which focus on different types of collocations and various kinds of text types, registers, and genres.

In vocabulary learning and teaching, the significant role of collocations should be emphasised. Seal (1991) regards collocations as a vital characteristic of vocabulary knowledge and states that what should be offered to students is collocational knowledge not individual words. EFL/ESL teachers should make their students aware of lexical collocations because such awareness has been considered an essential aspect of language learning (Brown, 1974; Coznett, 2000; Lewis, 2000). The present study provides empirical support to this view and also to Sinclair's (1997) idea about the importance of the idiom principle. This leads to the principle of language use which is pedagogically at the heart of teaching and testing of language competence and should be used for the design of appropriate teaching materials and for the methodology of classroom instruction.