The Language Proficiency Interview (LPI) is the speaking test component of the TOEIC^® test, open to those with TOEIC scorers of 730 or higher. It was initially developed at the request of Japan's Ministry of Ministry of International Trade and Industry by the Chauncey Group, a subsidiary of the Educational Testing Service (ETS) in the U.S. The first LPI tests were administered in 1980.
In the 20-30 minute interview, candidates reputedly demonstrate their overall spoken English ability, which is evaluated on a holistic six-level (0-5) proficiency scale. This rating scale is based on the same scale used by the Foreign Service Institute (FSI) Language School of the U.S. Army. Interviews are conducted by certified native English-speaking interviewers at TOEIC-designated sites across Japan six times per year at a cost of 13,000 yen. In Japan, between 1,200 and 1,400 people take the test each year. The TOEIC test and LPI are also offered in 31 other countries worldwide.
While the types of tasks covered in this criterion-referenced test vary with each of the six proficiency levels, the actual contents of the test do not. Interviewers are supposed to introduce topics based on the interviewee's background knowledge and interests. This approach is taken to determine what interviewees can and cannot do with the language in order to identify their highest sustainable level of speech performance.

Rating

All interviews are audiotaped and rated independently by two certified native English-speaking raters. A candidate's rating is determined by comparing speaking performance in the interview to the description of one of six numerical holistic levels on the LPI scale. Ratings are given at these levels between 0 and 5, or at "plus" levels (1+, 2+, 3+, 4+). The interviewer checks two or more levels systematically to identify the level for which the candidate consistently meets the requirements. A candidate who can sustain speech at the next higher level at least 60% of the time is given a plus level. Language competence is determined by the following three aspects: functions, context/content, and accuracy. The six functional levels are:

[ p. 2 ]

0 - candidate has no functional ability

1 - candidate can create with the language (i.e., does not rely on memorized language); can ask and answer questions; can participate in short conversations

2 - candidate can fully participate in casual conversations; can express facts, describe, give instructions, report, and provide narration about current, past, and future activities

3 - candidate can converse in formal and informal situations, resolve problem situations, deal with non-routine topics, provide explanations, describe in detail, offer supported opinions, and make hypotheses

4 - candidate can tailor language to fit the audience; can counsel, persuade, negotiate, and represent a point of view

5 - candidate functions in a manner equivalent to that of a well-educated native speaker (Educational Testing Service, 1999, p. 11)

A fuller description in TOEIC publicity materials of what a Level 2 speaker can do adds the following information: A Level 2 speaker can talk about "concrete topics such as own background, family, interests, work, travel, and current events...[and is] understandable to native speakers not used to dealing with foreigners; sometimes miscommunicates" (IIBC, 1997, p. 7).
In addition, each candidate's performance is rated according to the following criterion, which are defined in prose, for each of the six main levels: interactive comprehension, structural control, lexical control, delivery/fluency, social/cultural appropriateness, and communication strategies.

Test Contents

The three-part test begins with a warm-up — a simple conversation to help the candidate relax and to give an early indication of the candidate's level. The next section, which forms the core of the test, consists of level checks and probes. At least one role-play is also part of this section. Level checks are used to find the candidate's highest sustainable level of performance. When one level is achieved and sustained over several tasks, the interviewer probes to the next level. If that level is successfully demonstrated, the interview continues to the next higher level, until there is a breakdown in the candidate's ability to communicate. When that breakpoint is reached, the interviewer goes back to the previous level and continues with any remaining tasks. Once all of the tasks for that level are completed, the main part of the test is over. The test concludes with a wind-down, ending on a positive note to help restore the interviewee's confidence.
There are several specific tasks which the interviewer must obtain clear language samples for in order to determine an interviewee's level. For example, a Level 2 test asks the interviewee to demonstrate the following functions (IIBC, 1997):

1. engage in casual conversation
   
2. express facts
   
3. give instructions
   
4. describe
   
5. report
   
6. provide narration about current, past, and future activities
   

[ p. 3 ]

Test Features

The LPI test is a highly controlled instrument that is monitored closely by the Chauncey group at all levels, from test development to administration to evaluation. There are, of course, strengths and weakness to it.
Beginning with a positive attribute, the test itself unfolds naturally, since it is largely based on the candidates' knowledge and interests (particularly at levels 0+ through 2+). Thus, each interview is unique in terms of the topics covered. Thanks to this feature, the interviews (as far as tests are concerned) tend to be an enjoyable and relatively natural conversation experience for both the candidate and the interviewer.

"[The LPI] may not be suitable for less mature candidates.

One of the drawbacks of this test is because the tasks at Level 3 and higher are more abstract, conceptual, and opinion-based, they require not only a high level of language proficiency, but also an advanced degree of maturity and a richer background than younger candidates may have. Despite the fact that many university-age (and younger) candidates speak fluently, they do not have the capacity to answer at the level of sophistication that the upper levels require. Therefore, the test may not be suitable for less mature candidates.
Another point worth noting is that most people who take the LPI training program do not end up becoming interviewers for TOEIC — even those who become certified. Instead, they take away with them the LPI interviewer's manual and evaluation criterion and work as inhouse interviewers at their companies, evaluating employees' English skills for job placement, promotion, and overseas transfer. If they work at language schools, they evaluate students in their TOEIC and related courses. Since these inhouse tests may not be given according to TOEIC procedures, their administration and evaluation may be inconsistent and/or compromised. Both the candidates and the leaders in companies who use these scores to make career decisions for their employees should be aware of this potential inaccuracy.

Becoming a LPI Interviewer

To become a certified interviewer, applicants must first complete a three-day intensive training workshop (at a cost of ¥60,000), then submit accurate evaluations of 10 audio taped interviews (provided by TOEIC), and finally conduct and submit five ratable audio taped interviews. The success rate for aspiring interviewers is approximately 20%. Once certified, interviewers participate in an annual refresher workshop in which fellow raters meet, review tapes together, and discuss issues, problems, and questionable points that have come up over the year. Recertification occurs three years. For this, interviewers conduct and submit up to four taped interviews for evaluation.

[ p. 4 ]

Despite the challenges the training period and certification upkeep pose, becoming a certified LPI interviewer offers a good opportunity to learn how to be a flexible interviewer, how to ask meaningful questions which identify the interviewee's proficiency level, and to learn how to quickly evaluate the responses.

References

Educational Testing Service. (1999). TOEIC^® language proficiency interview manual. Princeton, NJ: Author.

Institute for International Business Communication (IIBC). (1997, May). TOEIC^® Newsletter 58, p. 7.

---

Newsletter: Topic IndexAuthor IndexTitle IndexDate Index
TEVAL SIG: Main Page Background Links Network Join

---

HTML: http://www.jalt.org/test/mac_3.htm   /   PDF: http://www.jalt.org/test/PDF/MacGregor3.pdf

[ p. 5 ]