Technology-Enhanced Accent Modification for International Teaching Assistants

Purpose

The technology-enhanced accent modification program (TEAM) addresses complaints from students that their learning and grades suffer because they cannot understand their international teaching assistants (ITAs). The lack of oral proficiency of many ITAs is recognized as a major problem at colleges and universities across the nation, where approximately 45,000 ITAs teach many of the 15 million students.

TEAM's creators sought to make accent modification instruction more accessible by developing reliable multimedia software that operates on low cost personal computers. The goal was not to teach ITAs to speak just like native English speakers but to help them to become intelligible to their students.

Innovative Features

On the assumption that the more ways the learner can sense a feature, the more easily he or she will learn and use it, TEAM provides multisensory computer instruction and feedback so that ITAs can see as well as hear their speech. TEAM's multimedia also enables learners to retrieve, display, and play model utterances of speech features. This is possible because recent advances in computers allow program designers to combine the visual images of a feature with audio recordings.

On the computer monitor the user sees model voiceprints showing the pitch and volume of each sound in sample phrases. Below the model, the user sees voiceprints of his or her own efforts with the same phrases. In this way, ITAs get visual input for aurally shaping their speech patterns to fit the models. The visual comparison helps them to eliminate monotonous delivery, choppy speech, inappropriate syllable stress, and the omission of certain sounds.

A criterion-referenced "Accent Survey" identifies the features of accents that need to be modified for ITAs to make themselves more easily understood. The curriculum software addresses these features, and a database contains 3,600 (1,800 male and 1,800 female) model utterances. An instructional component with online assistance helps ITAs to modify their accents' most distracting features.

The TEAM strategy differs from other approaches to accent modification in significant ways. Whereas most other pedagogies do not focus on general speech comprehensibility, TEAM addresses the prosodic (pitch, loudness, timing) features of accent as well as the pronunciation of consonants and vowels.

The curriculum software is designed to be used by pre-professional or graduate students who act as accent tutors to the ITAs, thus lowering delivery costs. Project staff train the tutors to use the design protocol and to adhere to TEAM instructional tactics. The software is designed to teach ITAs to assume responsibility for making and maintaining improvements in their accents, and it uses off-the-shelf personal computers that are widely available and affordable.

Evaluation

The software was field tested on novice computer users, and debugged to run reliably and dependably. Once a fail-proof version was created that could withstand 2,000 hours of use, testing was expanded to 128 students at Cleveland State, Kent State, and the University of Toledo.

Data were gathered on both the classroom performance of students taught by ITAs and the oral proficiency of the ITAs themselves before and at several intervals following instruction. TEAM students were compared with three groups of teaching assistants receiving other kinds of English instruction, especially the commonly used oral proficiency model.

Project Impact

The field test results were more positive than anticipated. They indicated that students enrolled in classes and laboratories with TEAM-taught ITAs as instructors performed better and obtained slightly higher grades than students enrolled in classes taught by non-TEAM ITAs.

Further, fewer students of TEAM ITAs switched to other same-time classes or laboratories. A majority of students (72 percent) in classes taught by TEAM ITAs reported that it took little or no effort to understand their instructors, compared to 49 percent of students in classes taught by non-TEAM instructors. Seventy-one percent of students taught by TEAM ITAs said that their instructors' speech did not affect their teaching performance, 92 percent described their instructors' speech as comprehensible, and 87 percent said that their own classroom performance was not affected by their instructors' speech.

TEAM-taught ITAs also performed better than non-TEAM-taught ITAs on other measures. On the "Accent Survey," they received significantly higher posttest scores on accent improvement. They took less time to make instructional improvements and, several re-tests after instruction, they retained their oral proficiency gains better than non-TEAM ITAs. (At these re-tests other ITAs regressed to pre-instruction levels of performance.) The frequency of mispronunciations was measured by the Sikorski Oral Proficiency Test. Even though the TEAM approach devoted much less time to the pronunciation features of speech than to the prosodic features, at retest it still significantly decreased ITAs pronunciation errors. However, at subsequent intervals, TEAM ITAs lost most of their pronunciation gains, compared to ITAs taught under the oral proficiency approach.

Lessons Learned

It took considerable time and cajoling to persuade institutions to retest ITAs for oral proficiency after remedial coursework was completed. The same resistance was encountered in getting English as a Second Language departments to use and follow the protocol, and several test sites had to be dropped because they would not comply. The project director concluded that, in academic settings, political and territorial factors are greater obstacles to change than technological ones.

This project demonstrates how advances in technology have allowed speech-language pathologists to work more effectively in accent modification. Extracting and displaying the acoustic properties of sound is something that technologies do accurately, quickly and reliably. An early assumption, that auditory modality alone was not enough to provide students with the clues they need to change pitch, loudness, and timing, proved true. Students can now see as well as hear their speech. They can see and hear the pitch line rise or fall. They can also see the increased loudness and duration of syllable stress and the word endings that they tend to omit. These features help them make their speech easier to understand.

The accent modification technologies turned out to be much more engaging for ITAs than drill and practice approaches. Learning and reading graphic displays of their own voiceprints and comparing them to those of the model voice actively focuses learners on their tasks and holds their attention.

In addition, TEAM turned out to be remarkably cost effective. It costs $264 to deliver instruction to a student using the TEAM approach, compared to $335 using the oral proficiency approach. When long term retention of gains in oral proficiency as well as student outcomes are factored in, the TEAM approach is approximately 50 percent less expensive.

Program Continuation and Recognition

In 1995, the project received the National Award for Excellence in Teaching English from the English Speaking Union of the United States. The judges rated the TEAM project superior in all categories of innovation, measurable effectiveness, and overall value.

In August 1997, the New York Times education section devoted several pages to new English language programs, highlighting the TEAM approach at Cleveland State. Consequently, the project director has opened a small company, Accent Services, USA, using TEAM software to teach non-native professionals in the community in addition to his university students.

Dissemination

Complimentary copies of the TEAM software were distributed to more than 350 colleges and universities with the largest enrollments of international students. Lawrence Erlbaum Associates, a major publisher, now markets TEAM at an educational price of $495. In addition to the field test sites, the software has been adopted for use at the University of Memphis, Michigan State University, the University of Wisconsin, and the University of Texas-Austin. Articles based on the TEAM project can be found in at least six speech-language journals, as well as in national higher education publications. In addition, project staff are participating in a FIPSE dissemination grant involving Johns Hopkins and three other universities.

Available Information

As noted earlier, the TEAM software is commercially available for higher education institutions. The project's final report to FIPSE is available on disc.

For further information, contact:

Arthur H. Schwartz
Department of Speech and Hearing
Cleveland State University
Euclid Avenue at East 24th Street
Cleveland, OH 44115
Telephone: 216-687-6990

[The City University of New York] [Table of Contents ] [V. General Education]

FIPSE Home