Modern Language Aptitude Test and Manual

Modern Language Aptitude Test and Manual (MLAT)

Carroll, J.B., & Sapon, S.M. (1959). San Antonio, TX: The Psychological

Corporation.

Linda Steinman and Monika Smith, OISE/UT

The Modern Language Aptitude Test (MLAT) was designed to measure the ability of

English speakers to learn a foreign language quickly in classroom settings. The test,

used for selection, placement, and diagnostic purposes, was developed over 40

years ago and has not been updated in any way since then. It is still in use, however,

and this intrigued us. In this paper, we take a critical look at the MLAT to determine

why this test has retained a firm hold in the area of predicting foreign language

aptitude and to explore whether it should, at this time, be updated.

The test

The MLAT measures four skills: phonetic coding ability, grammatical sensitivity,

memory ability, and inductive language learning ability. The names of the five test

sections and the tasks that the test- takers must perform are as follows:

?Part I: Number Learning (aural)

- Listen on tape to pseudo-words for numbers and

then transcribe numbers of up to three digits.

?Part II: Phonetic Script (aural)

- Learn and then demonstrate recognition of

correspondences between speech sounds and orthographic symbols.

?Part III: Spell

ing Clues - Read words that are spelled as they are pronounced and

choose synonyms for them.

?Part IV: Words in Sentences

- Note selected words in model sentences and locate

words that have similar functions in other sentences.

?Part V: Paired Associates

- Memorize a lexicon of 24 words from another language,

practise, and then be tested on those words.

Case study

In order to examine face validity and to determine first hand the ease of test

operation, we administered the MLAT to a volunteer graduate student. A native

English speaker, he was interested in knowing how proficient a language learner he

is. We selected the long form of the MLAT, which takes one hour to complete.

Operationally, the test is a pleasure. All that is needed is a tape recorder, the

pre-recorded MLAT cassette, a test booklet, an examinee answer sheet, a

test-scoring sheet, an administrator's manual, and a pencil. The manual is excellent

and answered most of our questions. The tape delivers all instructions to the

test-taker and even allows for time allotments.

We interviewed our volunteer after he wrote the test. Some of his comments follow:

'It was quite hard. There was no time to think about how you were doing. I am used

to essay-type questions, so the multiple choice was different for me and it was

sometimes hard to keep my place on the scoring sheet.' (We noticed his time-saving

strategy of not filling in the circles on his answer sheet with pencil; he just circled

them and went back after to 'colour them in.')

We asked, 'Would you say it is a fair test? Would you accept it as a fair predictor of

your foreign language aptitude?' He replied, 'Yes, the test was a good one. It was

long enough and it had a range of questions. But I know there is more to language

learning that what was on the test, such as motivation and personality.'

One of our expectations was that the volunteer would find the test vocabulary or the

test format dated. We asked, 'Did the test seem current?' He replied, 'The voice on

the tape was a little old, but otherwise it seemed fine.'

In our minds, both face validity and ease of test administration were established.

A point of interest: it is stated in the manual that while it is preferable to administer

the long form of the test, the short form (which takes 30 minutes and omits Parts I

and II) should yield similar results. This did not hold true for our volunteer, who did

far better on the first two parts of his test. If he had taken the short form, his score

would have been significantly lower.

Validity and concerns

The MLAT was developed over five years of extensive research into the prediction of

success in foreign languages. Experimental tests were administered to

approximately 5,000 people. Four relatively independent components of language

aptitude were identified for the final version of the test. The test was validated

thoroughly, and validation has been ongoing since then (Ehrman & Oxford, 1995;

Grigorenko, Sternberg, & Ehrman, 2000; Sparks, Ganschow, & Patton, 1995).

The MLAT is also the established benchmark test for validating other measures of

language aptitude or predictors of language proficiency. Some recent examples

include the VORD (Parry & Child, 1990) and the CANAL-FT (Grigorenko, Sternberg,

& Ehrman, 2000).

The MLAT has consistently achieved good results and has high predictive value, but

it has not gone unquestioned. A frequent challenge to the MLAT has been that it

does not reflect current knowledge in the field of cognitive psychology (Oxford,

1990; Rees, 2000; Stansfield, 1989). Researchers, however, find it difficult to agree

on how to define, and then on how to measure, many of these traits (Rychlak, 1981,

cited in Rees, 2000; Sternberg, 1995). There appear to be, for example,

irreconcilable differences in theories about what constitutes personality and

contention over distinguishing learning style from ability.

Does the MLAT measure language aptitude? We are satisfied that the test measures

the four components mentioned earlier, but Carroll himself agrees that not all

possible aspects of language aptitude are included in the MLAT (Carroll, 1990). His

model indicates that language proficiency results from a combination of aptitude,

motivation, institutional setting, and other possible factors. An aptitude test,

therefore, will predict success in a language course only to some degree. Aside from

the difficulties already mentioned in determining which other factors to include and

how to measure them reliably, expanding the aptitude test might generate

practicability problems. The test might become too long and/or harder to administer.

Furthermore, the predictive success of an aptitude test will itself depend on the

reliability of the achievement measurement to which it is compared (Goodman et al.,

1990; Green, 1975).

Another challenge to the MLAT has been the change in classroom methodology since

the test's development. Language training is now more communicative. Does this

make the MLAT less useful? Carroll reports in the MLAT manual that different

approaches to language teaching did not significantly alter the validity of the test

scores (p. 22). The MLAT continues to perform well even in recent studies (Ehrman

& Oxford, 1995), achieving good correlations with proficiency measures in modern

classrooms. Perhaps the MLAT measures metalinguistic awareness which is

independent of classroom methodology (Sparks, Ganschow, & Patton, 1995).

Changes in classroom methodology, therefore, do not seem to indicate a pressing

need for a change to or update of the test.

Need for updating

The MLAT manual provides norm tables for high school students (Grades 9-12) as

well as for adult learners. However, we had difficulties deciding on a suitable table

for evaluating our volunteer's performance. As an adult graduate student, he

seemed to fit neither the table for high school student nor the table for members of

(mostly) the Armed Forces enrolled in language schools. In addition, the norms

tables in the manual were established in 1958, when the high school population

would have been quite different from more recent cohorts. The importance of

establishing local norms for determining cut-off scores is emphasized by Ehrman

and Oxford (1995) and in the MLAT manual (p. 20). However, when, as in our case,

individuals (not groups) are tested, no comparison with other members of a cohort

is available. It would therefore be useful if an update of the manual were to include

some new norms tables.

Carroll (1990) mentions a need to 'fine-tune' the MLAT and suggests a possible

alternate form of the test to relieve test-giver tedium, prevent leakage, and permit

retesting of individuals. Minor defects cited by Carroll include a lack of clear notice

about the speeded nature of Part III. Our volunteer test-taker was not aware of the

short time available for answering this section and consequently ran out of time half

way through. We recommend that a more explicit warning be added to test

instructions on the answer sheet and on the tape. Similarly, in Part I, each digit of

the three-digit number is scored separately. This is not made clear to test-takers,

and they might well feel that it is not worth entering any of the digits if they mishear

one of them. Thus, they would lose the opportunity to earn partial marks.

We would like to see some minor changes to the manual itself, such as the deletion

of instructions regarding the old reel- to-reel tape recorders. The current version of

the test comes with a cassette tape, so the directions should correspond.

Finally, as the test is clearly still valuable (Ehrman & Oxford, 1995, recommend

using the MLAT more often), a computer version would appear appropriate.

We return to the question of the MLAT'S staying power and offer the following

possibilities. Firstly, a prodigious amount of testing has gone into the MLAT, and

Carroll has countered many of the challenges to the test. Secondly, many tests and

studies done in the last 40 years have used the MLAT as a benchmark. To call the

MLAT into question would call many subsequent studies into question, too.

There is, however, current activity in the development of foreign language aptitude

tests. Jonathan Rees (2000) writes about a test being developed now at the

University of Birmingham, where he believes that 'the time is ripe, both

intellectually and commercially' for new research in this area. As well, the CANAL-F

Test (Cognitive Ability for Novelty in Language Acquisition-Foreign) is currently

being refined.

For now, the MLAT appears to serve its purpose. With some updating, it is likely to

continue to be a strong test instrument for many years to come.

References

Carroll, J.B. (1990). Cognitive abilities in foreign language aptitude: Then and now.

In T.S. Parry & C.W. Stansfield (Eds.), Language aptitude reconsidered (pp. 11-29).

Englewood Cliffs, NJ: Prentice Hall Regents.

Ehrman, M.E., & Oxford, R.L. (1995). Cognition plus: Correlates of language

learning success. Modern Language Journal, 79, 67-89.

Goodman, J.F., et al. (1990). Determining exemptions from foreign language

requirements: Use of the Modern Language Aptitude Test. Contemporary

Educational Psychology, 15, 131-141.

Green, P. (1975). Aptitude testing: An on-going experiment. Audio- Visual Language

Journal, 12, 205-210.

Grigorenko, E.L., Sternberg, R.J., & Ehrman, M.E. (2000). A theory- based approach

to the measurement of foreign language learning ability: The Canal-F theory and

test. Modern Language Journal, 84, 390-405.

Oxford, R. (1990). Styles, strategies and aptitude: Connections for language

learning. In T.S. Parry & C.W. Stansfield (Eds.), Language aptitude reconsidered

(pp. 67-125). Englewood Cliffs, NJ: Prentice Hall Regents.

Parry, T.S., & Child, J.R. (1990). Preliminary investigation of the relationship

between VORD, MLAT, and language proficiency. In T.S. Parry & C.W. Stansfield

(Eds.), Language aptitude reconsidered (pp. 30-66). Englewood Cliffs, NJ: Prentice

Hall Regents.

Rees, J. (2000). Predicting the future of foreign language aptitude testing. In S.

Cornwell & P. Robinson (Eds.), Individual differences in foreign language learning:

Effects of aptitude, intelligence, and motivation. Conference proceedings, Aoyama

Gakuin University, Tokyo.

Sparks, R.L., Ganschow, L., & Patton, J. (1995). Prediction of performance

in first-year foreign language courses: Connections between native

and foreign language learning. Journal of Educational Psychology, 87, 638-655.

Spolsky, B. (1995). Prognostication and language aptitude testing. Language

Testing, 12, 321-340.

Stansfield, C.W. (1989). What is foreign language aptitude? Washington, DC: Eric

Clearinghouse. (ERIC Documentation Reproduction Service No. ED 318 226)

Sternberg, R. (1995). Styles of thinking and learning. Language Testing, 12,

259-291.

? 2001 The Canadian Modern Language Review/La Revue canadienne des langues

vivantes, 58, 2 (December/dé

cembre)

The College of Oxford University

“

Mode N”

Test for Entrance to Classics 1995

LANGUAGE APTITUDE TEST

Time allowed: 60 minutes

Please write your name and college of first choice at the top of this page.