Literature Review of Language Testing Theories and Approaches

Abstract

Language testing is always considered as a part of language learning and teaching. This paper reviews the historical trends of language testing theories and testing approaches according to Spolsky’s opinion. Due to the close relations between language testing and linguistics as well as language teaching, this review also deals with the corresponding linguistic background. In conclusion, the paper points out the connections of language testing with other disciplines, and the phenomenon of coexistence of several different testing theories and approaches.

Share and Cite:

Mao, A.M. (2022) Literature Review of Language Testing Theories and Approaches. Open Access Library Journal, 9, 1-5. doi: 10.4236/oalib.1108741.

1. Introduction

As a branch of applied linguistics, language testing is a relatively new discipline with a history of no more than 100 years. The development of language testing theory is closely related to people’s views of language and language use. Moreover, language testing is inseparable from language learning and teaching, no matter which kind of testing theory is taken. For language teachers, tests perform both pedagogical and research functions.

This essay is a brief literature review of the developments of language testing theories and corresponding testing approaches. In the latter part, the essay presents some prevailing views concerning language testing.

2. Language Testing Theories and Approaches

Bernard Spolsky (1978) [1] distinguishes three historical periods of modern language testing: the pre-scientific period, the psychometric-structuralist period and the integrative-sociolinguistic (or the psycholinguistic-sociolinguistic) period.

Its development has always been keeping pace with that of linguistics (from historic comparative linguistics to structuralist linguistics, and finally to psycholinguistics and sociolinguistics) and that of language teaching methodology (from grammar-translation method to audio-lingual method, and last to communicative approach). The following part focuses on the three testing theories on the base of contemporary linguistics.

2.1. Pre-Scientific Testing (Prior to the Early 1950s)

Language testing before the early 1950s was collectively called “the pre-scientific testing”. Language was not scientifically defined over that period and was simply taught as “knowledge”, consisting of grammatical knowledge, lexical knowledge and phonetic knowledge. Accordingly, language testing then was only to test knowledge in these three aspects.

The test focused on specific language points selected by the teacher from the textbook, and its main content was grammatical rules, morphological change and the usage of words. In a paper-and-pencil format, the test paper was always designed and scored by the teacher of this course, without any standard specifications or scoring criteria. Consequently, the test always relied on the personal subjective analytic ability of that teacher and required no special techniques for testing. As for the item types, most of them were subjective items―translation, writing, grammar analysis and question-and-answer, dealing with the knowledge of literature and culture. In short, pre-scientific testing was empirical and subjective, without a theoretical base. It can be seen as a preliminary step toward the current scientific tests.

2.2. Psychometric-Structuralist Testing (The Early 1950s to the Late 1960s)

Influenced by structuralist linguistics, language then was defined as a system of symbols. It was composed of independent components (grammar, vocabulary and pronunciation) and independent skills (listening, speaking, reading and writing). Together with behaviorist psychology, structural linguistics brought in a new teaching method, namely, the Audio-lingual Method (Castagnaro, 2006) [2].

Due to this popular view of language, there was “a tendency to atomize and decontextualize the knowledge to be tested and to test aspects of knowledge in isolation”. (McNamara, 2003) [3] With the application of psychometrics into language testing, a new approach of testing, known as discrete point testing, appeared. In accordance with the discrete point testing, there must be only one testing point in each item, and the tests of grammar would be separated from other language components. Moreover, materials in the tests were presented with minimal context. To suit these characteristics, multiple-choice was chosen as the main item type for the testing and there was always an adequate sampling of the items to achieve validity. Besides, there were other objective items, like sentence completion, blank-filling and error correction.

2.3. Psycholinguistic-Sociolinguistic Testing (The Late 1960s to the Present Time)

2.3.1. Integrative Test

During the 1960s, discrete point testing was a thriving testing approach. However, with the coming of the “integrative-sociolinguistic” era with its emphasis on communication, authenticity and context, the discrete point approach obviously met with some criticism. It was seen as “focusing too exclusively on knowledge of the formal linguistic system for its own sake rather than on the way such knowledge is used to achieve communication” (Parviz & Rajab, 2009) [4]. Gradually, tests tended to integrate different components of language (pronunciation, grammar, vocabulary) within a specific context.

John Oller is an important figure who pushed forward the integrative test and succeeded in making it less expensive and easier to score. Oller (1979) [5] argued that language competence is a unified set of interacting ability that cannot be separated apart and tested adequately and thus he proposed Unitary Competence Hypothesis, which demonstrated that tests would pay attention to candidates’ general language proficiency. Accordingly, Oller (1979) [5] strongly recommended cloze tests for the test results are good measures of overall language proficiency (reading comprehension, grammar, and vocabulary as well as background knowledge). In addition to the cloze test, dictation was also held up as the prime item type for integrative tests.

Unfortunately, further work soon showed drawbacks of the cloze test and dictation, which were called “pragmatic tests” by Oller (1979) [5]. Scholars against them claimed that dictation failed to directly and correctly measure whether the test takers could finish tasks in a real situation―dictation was radically indirect and thus lacked validity (Harris, 1969) [6]. Also, the cloze test was criticized as a combination of discrete point tests of grammar and vocabulary.

2.3.2. Communicative Language Test

Both the integrative test and the communicative test emphasize the significance of language (use of language) instead of the form and structure of language (usage of language). However, the latter attaches more importance to the use of all kinds of language skills, not two or more than two skills as required by the integrative test.

In the 1960s, Noam Chomsky (1965) [7] proposed two concepts―“linguistic competence” and “linguistic performance”, which refreshed people’s view of language and language use. Later, in the early 1970s, Hymes’s (1972) [8] theory of communicative theory about communicative competence began to exert a significant influence on language teaching and potentially testing. Hymes (1972) [8] saw that knowing language was more than knowing its rules of grammar. There were culturally specific rules which related the language used to features of the communicative context. Based on this, communicative language tests came to have two features (McNamara, 2003) [3]: 1) they were performance tests, requiring the candidate to be engaged in a specific communication context; 2) they paid attention to the social roles that candidates were likely to be in real-world settings, and offered specified demands of such roles. Here, the focus on test takers’ social roles exemplified a shift from a psychological perspective on language to a sociological one.

In the early 1980s, Michael Canale and Merrill Swain (1980) [9] specified the components of “communicative competence” as the following four aspects: grammatical competence, sociolinguistic competence, strategic competence and discourse competence.

In the 1990s, another new model of “communicative competence” prevailed. This model, proposed by Bachman (1999) [10], consists of two parts―communicative language ability (CLA) and test method facets (TMF). The following is a brief introduction to Bachman’s “interactional approach to language testing”, which was praised as a milestone in the history of language testing by Skehan (1991) [11].

In the framework of CLA, three parts were contained―language competence, strategic competence and psychophysiological mechanisms. Bachman (1999) [10] further divided language competence into organizational competence and pragmatic competence. The former (organizational competence) refers to one’s ability to make grammatically correct sentences, understand these sentences and arrange them into a logical passage, and can be subdivided into grammatical competence and textual competence. The latter (pragmatic competence) emphasizes how to speak in a specific context, and also can be subdivided into semantic competence, function competence and sociolinguistic competence. As for strategic competence, Bachman (1999) [10] was dissatisfied with the previous one proposed by Canale and Swain (1980) [9]. In his view, strategic competence consisted of four inseparable metacognitive strategies―assessment strategies, goal-setting strategies, planning strategies and execution strategies.

In a word, the framework of CLA deals with what needs to be tested in communicative testing. As for the specific method of constructing communicative language tests, Bachman (1999) [10] put forward the TMF, which includes five facets: the testing environment, the test rubric, the nature of the input the test taker receives, the nature of the expected response to that input, the relationship between input response.

To summarize, there are several characteristics of a communicative language test: 1) it is context-specific according to candidates’ needs; 2) it tests language in a broader way, including language knowledge, its functions and its appropriateness; 3) it adopts the qualitative modes of assessment; 4) the tasks are dependent, i.e. one task is based on the previous one; 5) it requires candidates to get underlying information from given information.

3. Conclusion

Language testing is commonplace in our language learning process. Various kinds of language tests (e.g. aptitude tests, diagnostic tests, placement tests, achievement tests, and proficiency tests) exist both at home and abroad.

Language testing cannot be isolated from language learning and teaching. Throughout the history of language testing, it is evident that testing theory and approach always follow the relevant views of language and teaching methodology. The “backwash” effect of testing on teaching is worth the attention, especially from teachers. Besides, the development of language testing also needs the knowledge of other subjects, such as statistics and metrology.

The same as in other fields (e.g. teaching methodology), there is a phenomenon of coexistence of several different testing theories or approaches. The combination of discrete-point testing and integrative testing is very common in a test, such as TEM-4 with both dictation, cloze (integrative) and multiple-choice (discrete). It is hard to judge which item type is good and which is bad. What is essential is whether these items fit into the test purpose and are suitable for the test-takers.

Conflicts of Interest

The author declares no conflicts of interest.

References

[1] Spolsky, B. (1978) Educational Linguistics: An Introduction. Newbury House Publishers, Rowley, MA.
[2] Castagnaro, P.J. (2006) Audiolingual Method and Behaviorism: From Misunderstanding to Myth. Applied Linguistics, 3, 519-526. https://doi.org/10.1093/applin/aml023
[3] McNamara, T. (2003) Language Testing. Shanghai Foreign Language Education Press, Shanghai.
[4] Parviz, A. and Rajab, E. (2009) A Close Look at the Relationship between Multiple Choice Vocabulary Test and Integrative Cloze Test of Lexical Words in Iranian Context. English Language Teaching, 2, 163-170. https://doi.org/10.5539/elt.v2n3p163
[5] Oller, J. (1979) Language Tests at School: A Pragmatic Approach. Longman, London.
[6] Harris, D.P. (1969) Testing English as a Second Language. McGraw-Hill, New York.
[7] Chomsky, N. (1965) Aspects of the Theory of Syntax. The M.I.T. Press, Cambridge, MA. https://doi.org/10.21236/AD0616323
[8] Hymes, D. (1972) On Communicative Competence. In: Pride, J.B. and Holmes, A., Eds., Sociolinguistics: Selected Readings, Penguin, Harmondsworth, 269-293.
[9] Canale, M. and Swain, M. (1980) Theoretical Bases of Communicative Approaches to Second Language Teaching and Testing. Applied Linguistics, 1, 1-47. https://doi.org/10.1093/applin/1.1.1
[10] Bachman, L.F. (1999) Fundamental Considerations in Language Testing. Shanghai Foreign Language Education Press, Shanghai.
[11] Skehan, P. (1991) Individual Differences in Second Language Learning. Studies in Second Language Acquisition, 13, 275-298. https://doi.org/10.1017/S0272263100009979

Copyright © 2024 by authors and Scientific Research Publishing Inc.

Creative Commons License

This work and the related PDF file are licensed under a Creative Commons Attribution 4.0 International License.