TEST&SCORE

MANUAL

1997 Edition

®

T he TOEFL Test and Score Manual has been prepared for deans,admissions officers and graduate department faculty, adminis-trators of scholarship programs, ESL teachers, and foreign studentadvisers. The Manual provides information about the interpretationof TOEFL® scores, describes the test format, explains the operationof the testing program, and discusses program research activities andnew testing developments.

This edition of the Test and Score Manual updates material inthe 1995-96 edition, providing a description of revisions to the testintroduced in July 1995, and other information of interest to scoreusers. With the exception of “Program Developments” on page 10,information in this Manual refers specifically to the paper-and-pencilTOEFL test.

As this edition goes to press (summer 1997), a computer-based TOEFL test is under development and planned forintroduction in the second half of 1998 (see page 11). Moreinformation about the computer-based test and score interpreta-tion will appear on the TOEFL website at http://www.toefl.organd through new publications as it becomes available.

Add your name to the TOEFL web list (on “Educators”directory page) and receive e-mail announcements as they arereleased.

TOEFL Programs and ServicesInternational Language ProgramsEducational Testing Service

Educational Testing Service is an Equal Opportunity/Affirmative Action Employer.

Copyright © 1997 by Educational Testing Service. All rights reserved.

EDUCATIONAL TESTING SERVICE, ETS, the ETS logo, GRADUATE RECORD EXAMINA-TIONS, GRE, SLEP, SPEAK, THE PRAXIS SERIES: PROFESSIONAL ASSESSMENTS FORBEGINNING TEACHERS, TOEFL, the TOEFL logo, TSE, and TWE are registered trademarksof Educational Testing Service.

COLLEGE BOARD and SAT are registered trademarks of the College Entrance Examination Board.

GMAT and GRADUATE MANAGEMENT ADMISSION TEST are registered trademarks of theGraduate Management Admission Council.

SECONDARY SCHOOL ADMISSION TEST and SSAT are registered trademarks of the SecondarySchool Admission Test Board.

No part of this publication may be reproduced or transmitted in any form or by any means,electronic or mechanical, including photocopy, recording, or any information storage and retrievalsystem, without permission in writing from the publisher. Violators will be prosecuted in accordancewith both United States and international copyright and trademark laws.

Permissions requests may be made on-line at http://www.toefl.org/copyrigh.html or sent to:Proprietary Rights OfficeEducational Testing ServiceRosedale RoadPrinceton, NJ 08541-0001USA

Phone: 609-734-5032

(continued)

CONTENTS ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7TOEFL Policy Council . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Committee of Examiners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Finance Committee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Research Committee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Outreach and Services Committee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9TWE Test (Test of Written English) Committee . . . . . . . . . . . . . . . . . . . . 9TSE Test (Test of Spoken English) Committee . . . . . . . . . . . . . . . . . . . . . 9

Program Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10TOEFL 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10The Computer-Based TOEFL Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Test of English as a Foreign Language . . . . . . . . . . . . . . . . . . . . . . . 11Use of Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Description of the Paper-Based TOEFL Test . . . . . . . . . . . . . . . . . . . . . . . 11Development of TOEFL Test Questions . . . . . . . . . . . . . . . . . . . . . . . . . . 12

TOEFL Testing Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13Friday and Saturday Testing Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . 13TWE Test (Test of Written English) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13TSE Test (Test of Spoken English) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13Institutional Testing Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Procedures at Test Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Measures to Protect Test Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Identification Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Photo File Records . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Photo Score Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Checking Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Supervision of Examinees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Preventing Access to Test Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

TOEFL Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Release of Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Test Score Data Retention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Image Score Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Official Score Reports from ETS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Information Printed on the Official Score Report . . . . . . . . . . . . . . . . . . . 20Examinee Score Records . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Acceptance of Test Results Not Received from ETS . . . . . . . . . . . . . . . 21Additional Score Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Confidentiality of TOEFL Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Calculation of TOEFL Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Hand-Scoring Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Scores of Questionable Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Examinees with Disabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Use of TOEFL Test Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3

4

Statistical Characteristics of the Test . . . . . . . . . . . . . . . . . . . . . . . . . 29Level of Difficulty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Test Equating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Adequacy of Time Allowed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Reliabilities and the Standard Error of Measurement . . . . . . . . . . . . . . . . 29Reliability of Gain Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32Intercorrelations Among Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

Content Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Criterion-Related Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Construct Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

Other TOEFL Programs and Services . . . . . . . . . . . . . . . . . . . . . . . . . 39TWE Test (Test of Written English) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39TSE Test (Test of Spoken English) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39SPEAK Kit (Speaking Proficiency English Assessment Kit) . . . . . . . . . . 40SLEP Test (Secondary Level English Proficiency Test) . . . . . . . . . . . . . . 40Fee Voucher Service for TOEFL and TSE Score Users . . . . . . . . . . . . . . . 40TOEFL Fee Certificate Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41TOEFL Magnetic Score-Reporting Service . . . . . . . . . . . . . . . . . . . . . . . . 41Examinee Identification Service for TOEFL and TSE Score Users . . . . . 41Support for External Research Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

Research Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43TOEFL Research Report Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43TOEFL Technical Report Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45TOEFL Monograph Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47TOEFL Products and Services Catalog . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Bulletin of Information for TOEFL, TWE, and TSE . . . . . . . . . . . . . . . . 47Test Center Reference List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Test Forms Available to TOEFL Examinees . . . . . . . . . . . . . . . . . . . . . . . 47Guidelines for TOEFL Institutional Validity Studies . . . . . . . . . . . . . . . . 47TOEFL Test and Score Data Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Institutional Testing Program Brochure . . . . . . . . . . . . . . . . . . . . . . . . . . . 48TOEFL Test of Written English Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . 48TSE Score User’s Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48Secondary Level English Proficiency Test Brochure . . . . . . . . . . . . . . . . . 48The Researcher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

TOEFL Study Materials for the Paper-Based Testing Program . 49

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

ETS Offices Serving TOEFL Candidates and Score Users . . . . . . 54

TOEFL Representatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

5

TABLES ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

Table 1. Minimum and Maximum Observed Section and Total Scores . . 23

Table 2. Reliabilities and Standard Errors of Measurement (SEM) . . . . . 30

Table 3. Intercorrelations Among the Scores . . . . . . . . . . . . . . . . . . . . . . . 33

Table 4. Correlations of Total TOEFL Scores with University Ratings . . 35

Table 5. Correlations of TOEFL Subscores with Interview andEssay Ratings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

Table 6. TOEFL/GRE Verbal Score Comparisons . . . . . . . . . . . . . . . . . . . 37

Table 7. TOEFL/SAT and TSWE Score Comparisons . . . . . . . . . . . . . . . 37

Table 8. TOEFL, GRE, and GMAT Score Comparisons . . . . . . . . . . . . . . 37

Table 9. Correlations Between GMAT and TOEFL Scores . . . . . . . . . . . . 38

6

7

OVERVIEW ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

The purpose of the Test of English as a ForeignLanguage (TOEFL®) is to evaluate the Englishproficiency of people whose native language is notEnglish. The test was initially developed to measurethe English proficiency of international studentswishing to study at colleges and universities in theUnited States and Canada, and this continues to beits primary function. However, a number of academicinstitutions in other countries, as well as certainindependent organizations, agencies, and foreigngovernments, have also found the test scores useful.The TOEFL test is recommended for students at theeleventh-grade level or above; the test content isconsidered too difficult for younger students.

The TOEFL test was developed for use starting in1963-64 through the cooperative effort of more than30 organizations, public and private. A NationalCouncil on the Testing of English as a ForeignLanguage was formed, composed of representativesof private organizations and government agenciesconcerned with testing the English proficiency offoreign nonnative speakers of English who wished tostudy at colleges and universities in the United States.The program was financed by grants from the Fordand Danforth Foundations and was, at first, attachedadministratively to the Modern Language Associa-tion. In 1965, the College Board® and EducationalTesting Service (ETS®) assumed joint responsibilityfor the program.

In recognition of the fact that many who take theTOEFL test are potential graduate students, a coop-erative arrangement for the operation of the programwas entered into by Educational Testing Service,the College Entrance Examination Board, and theGraduate Record Examinations® (GRE®) Board in1973. Under this arrangement, ETS is responsiblefor administering the TOEFL program accordingto policies determined by the TOEFL Policy Council.

Educational Testing Service. ETS is a non-profit organization committed to the developmentand administration of responsible testing programs,the creation of advisory and instructional services,and research on techniques and uses of measurement,human learning and behavior, and educationaldevelopment and policy formation. It develops andadministers tests, registers examinees, and operatestest centers for various sponsors. ETS also suppliesrelated services; e.g., it scores tests; records, stores,

and reports test results; performs validity studiesand other statistical studies; and undertakes programresearch. All ETS activities are governed by a16-member board of trustees composed of personsfrom the fields of education and public service.

In addition to the Test of English as a ForeignLanguage and the Graduate Record Examinations,ETS develops and administers a number of othertests, including the Graduate Management AdmissionTest®, and The Praxis Series: Professional Assess-ments for Beginning Teachers® tests, as well asthe College Board testing programs.

The Chauncey Group International Ltd., awholly-owned subsidiary of ETS, provides assess-ment, training, and guidance products and servicesin the workplace, military, professional, and adulteducational environments.

College Board. The College Board is a nonprofit,educational organization with a membership of morethan 2,800 colleges and universities, schools, andeducational associations and agencies. The CollegeBoard’s board of trustees is elected from the member-ship, and institutional representatives serve onadvisory councils and committees that review theprograms of the College Board and participate inthe determination of its policies and activities.

The College Board sponsors tests, publications,software, and professional conferences and trainingin the areas of guidance, admissions, financial aid,credit by examination, and curriculum improvementin order to increase student access to higher educa-tion. It also supports and publishes research studiesabout tests and measurement and conducts studieson education policy developments, financial aid needassessment, admissions planning, and related educa-tion management topics.

One major College Board service, the SAT® Pro-gram, includes the SAT I: Reasoning Test, and SAT II:Subject Tests. Subject Tests are available in suchdiverse content areas as writing, literature, languages,math, sciences, and history. The College Board con-tracts with ETS to develop these tests, operate testcenters in the United States and other countries, scorethe answer sheets, and send score reports to examineesand to the institutions they designate as recipients.

8

Graduate Record Examinations Board.The GRE Board is an independent board affiliatedwith the Association of Graduate Schools and theCouncil of Graduate Schools in the United Statesand Canada. It is composed of 18 representativesof the graduate community. Standing committeesof the board include the Research Committee, theServices Committee, and the Minority GraduateEducation Committee.

ETS carries out the policies of the GRE Boardand, under the auspices of the board, administersand operates the GRE program. Two types of testsare offered: a General Test and Subject Tests in 16disciplines. ETS develops the tests, maintains testcenters in the United States and other countries,scores the answer sheets, and sends score reports tothe examinees and to the accredited institutions andapproved fellowship sponsors the examinees designateas recipients. ETS also provides information, technicaladvice, and professional counsel, and develops propos-als to achieve the goals formulated by the board.

In addition to its tests, the GRE program offersmany services to graduate institutions and to prospec-tive graduate students. Services to institutions includeresearch, publications, and advisory services to assistgraduate schools and departments in admissions,guidance, placement, and the selection of fellowshiprecipients. Services to students include test familiar-ization materials and services related to informingstudents about graduate education.

TOEFL Policy CouncilPolicies governing the TOEFL program are formu-

lated by the 15-member TOEFL Policy Council. TheCollege Board and the GRE Board each appoint threemembers to the Council. These six members comprisethe Executive Committee and elect the remainingnine members. Some of these members-at-large areaffiliated with such institutions and agencies asgraduate schools, junior and community colleges,nonprofit educational exchange organizations, andother public and private agencies with interest ininternational education. Others are specialists in thefield of English as a foreign or second language.

There are six standing committees of the Council,each responsible for specific areas of program activity.

Committee of Examiners

The TOEFL Committee of Examiners is composedof seven specialists in linguistics, language testing, orthe teaching of English as a foreign or second language.Members are rotated on a regular basis to ensure thecontinued introduction of new ideas and philosophiesrelated to second language teaching and testing.

The primary responsibility of this committee is toestablish overall guidelines for the test content, thusassuring that the TOEFL test is a valid measureof English language proficiency reflecting currenttrends and methodologies in the field. The committeedetermines the skills to be tested, the kinds of ques-tions to be asked, and the appropriateness of thetest in terms of subject matter and cultural content.Committee members review and approve the policiesand specifications that govern the test content.

The Committee of Examiners not only lends itsown expertise to the test and the test developmentprocess but also makes suggestions for research and,on occasion, invites the collaboration of other authori-ties in the field, through invitational conferences andother activities, to contribute to the improvement ofthe test. The committee works with ETS test develop-ment specialists in the actual developmentand review of test materials.

Finance Committee

The TOEFL Finance Committee consists of at leastfour members and is responsible to the TOEFLExecutive Committee. The members develop fiscalguidelines, monitor and review budgets, and providefinancial analysis for the program.

9

Research Committee

An ongoing program of research related to theTOEFL program of tests is carried out under thedirection of the Research Committee. Its six membersinclude representatives of the Policy Council and theCommittee of Examiners, as well as specialists fromthe academic community. The committee reviewsand approves proposals for test-related research andsets guidelines for the entire scope of the TOEFLresearch program.

Because the studies involved are specific to theTOEFL testing programs, most of the actual researchwork is conducted by ETS staff members rather thanby outside researchers. However, many projectsrequire the cooperation of consultants and otherinstitutions, particularly those with programs in theteaching of English as a foreign or second language.Representatives of such programs who are interestedin participating in or conducting TOEFL-relatedresearch are invited to contact the TOEFL office.

As research studies are completed, reports arepublished and made available to anyone interested inthe TOEFL tests. A list of those in print at the timethis Manual was published appears on pages 38-40.

Outreach and Services Committee

This six-member committee is responsible forreviewing and making recommendations to improveand modify existing program outreach activities andservices, especially as they relate to access and equityconcerns; initiating proposals for the development ofnew program products and services; monitoring theCouncil bylaws; and carrying out additional tasksrequested by the Executive Committee or the Council.

TWE® Test (Test of Written English)Committee

This seven-member group consists of writing andESL composition specialists with expertise in writingassessment and pedagogy.

The TWE Committee, with ETS test developmentspecialists, is responsible for developing, reviewing,and approving test items for the TWE test. Thecommittee also prepares item writer guidelines andmay suggest research or make recommendations forimproving the TWE test to ensure that the test is avalid measure of English writing proficiency.

TSE® Test (Test of Spoken English)Committee

This committee has six members who have exper-tise in oral proficiency assessment and represent theTSE constituency.

The TSE Committee, with ETS test developmentspecialists and program staff, oversees the TSE testcontent and scoring specifications, reviews test itemsand scoring procedures, and may make recommenda-tions for research or test revisions to assure that thetest is a valid measure of general speaking proficiency.

10

PROGRAM DEVELOPMENTS ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

TOEFL 2000The TOEFL 2000 project is a broad effort underwhich language testing at ETS will evolve into thetwenty-first century. The impetus for TOEFL 2000came from the various constituencies, includingTOEFL committees and score users. These groupshave called for a new TOEFL test that (1) is morereflective of models of communicative competence;(2) includes more constructed-response itemsand direct measures of writing and speaking;(3) includes test tasks integrated across modalities;and (4) provides more information than currentTOEFL scores about international students’ abilityto use English in an academic environment.

Changes to TOEFL introduced in 1995 (i.e.,eliminating single-statement listening comprehen-sion items, expanding the number of academiclectures and longer dialogs, and embedding vocabu-lary in reading comprehension passages) repre-sented the first step toward a more integrativeapproach to language testing. The next major stepwill be the introduction of a computer-basedTOEFL test in 1998. (See next column.)

TOEFL 2000 now continues with efforts thatwill lead to the next generation of computerizedTOEFL tests. These include:

n the development of a conceptual frameworkthat— takes into account models of communicative

competence— identifies various task characteristics

and how these will be used in theconstruction of language tasks

— specifies a set of variables associated witheach of these task components

n a research agenda that informs and supportsthis emerging framework

n a better understanding of the kinds ofinformation test users need and want from theTOEFL test

n a better understanding of the technologicalcapabilities for delivery of the TOEFL test intothe next century

A series of TOEFL 2000 reports that are part ofthe foundation of the project are now available (seepage 44). As future projects are completed, mono-graphs will be released to the public in this newresearch publication series.

The Computer-Based TOEFL TestTesting on computer is an important advancement thatenables the TOEFL program to take advantage of newforms of assessment made possible by the computerplatform. This reflects ETS’s commitment to create animproved English-language proficiency test that will

n better reflect the way in which peoplecommunicate effectively

n include more performance-based tasksn provide more information than the current TOEFL

test about the ability of international students touse English in an academic setting

The computer-based test is not just the paper testreformatted for the computer. While some questionswill be similar to those on the current test, others willbe quite different. For example, the Listening Compre-hension and Reading Comprehension sections willinclude new question types designed specifically forthe computer. In addition, the test will include an essaythat can be handwritten or typed on the computer. Theessay will measure an examinee’s ability to generateand organize ideas and support those ideas using theconventions of standard written English.

Some sections of the test will be computer-adaptive.In computer-adaptive testing (CAT), the computerselects a unique set of test questions based on the testdesign and the test taker’s ability level. Questions arechosen from a very large pool categorized by itemcontent and difficulty. The test design ensures fairnessbecause all examinees receive the same

n number of test questionsn amount of time (if they need it)n directionsn question typesn distribution of content

The CAT begins with a question of medium diffi-culty. The next question is one that best fits theexaminee’s performance and the design of the test. Thecomputer is programmed to make continuous adjust-ments in order to present questions of appropriatedifficulty to test takers of all ability levels.

The TOEFL program has taken steps to assure thatan individual’s test performance is not influenced by alack of computer experience. A computerized tutorial,designed especially for nonnative speakers of English,has been developed to teach the skills needed to takeTOEFL on computer.

For periodic updates on the computer-based TOEFLtest, visit TOEFL OnLine at http://www.toefl.org.

11

Use of ScoresThe TOEFL program encourages use of the test

scores by an institution or organization to help makevalid decisions concerning English language profi-ciency in terms of its own requirements. However,the institution or organization itself must determinewhether the TOEFL test is appropriate, with respectto both the language skills it measures and its level ofdifficulty, and must establish its own levels of accept-able performance on the test. General guidelines forusing TOEFL scores are given on pages 26-28.

TOEFL score users are invited to consult with theTOEFL program staff about their current or intendeduses of the test results. The TOEFL office will assistinstitutions and organizations contemplating use ofthe test by providing information about its applicabil-ity and validity in particular situations. It also willinvestigate complaints or information obtained aboutquestionable interpretation or use of reported TOEFLtest scores.

Description of the Paper-BasedTOEFL Test

The TOEFL test originally contained five sections.As a result of extensive research (Pike, 1979; Pitcherand Ra, 1967; Swineford, 1971; Test of English as aForeign Language: Interpretive Information, 1970), athree-section test was developed and introduced in1976. In July 1995, the test item format was modifiedsomewhat within the same three-section structureof the test.

Each form of the current (1997) TOEFL testconsists of three separately timed sections deliveredin a paper-and-pencil format; the questions in eachsection are multiple-choice, with four possibleanswers or options per question. All responses aregridded on answer sheets that are computer scored.

The total test time is approximately two and one-half hours; however, approximately three and one-halfhours are needed for a test administration to admitexaminees to the testing room, to allow them to enteridentifying information on their answer sheets, andto distribute and collect the test materials. Briefdescriptions of the three sections of the test follow.

n Section 1, Listening Comprehension

Section 1 measures the ability to understand Englishas it is spoken in North America. The oral featuresof the language are stressed, and the problems testedinclude vocabulary and idiomatic expression aswell as special grammatical constructions that arefrequently used in spoken English. The stimulusmaterial and oral questions are recorded in standardNorth American English; the response options areprinted in the test books.

There are three parts in the Listening Comprehen-sion section, each of which contains a specific typeof comprehension task. The first part consists of anumber of short conversations between two speakers,each followed by a single spoken question. Theexaminee must choose the best response to thequestion about the conversation from the four optionsprinted in the test book. In the second and third partsof this section, the examinee hears conversations andshort talks of up to two minutes in length. Theconversations and talks are about a variety of sub-jects, and the factual content is general in nature.After each conversation or talk the examinee is askedseveral questions about what was heard and, for each,must choose the one best answer from the choicesin the test book. Questions for all parts are spokenonly one time.

○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

TEST OF ENGLISH ASA FOREIGN LANGUAGE:The Paper-Based Testing Program

11

12

n Section 2, Structure and WrittenExpression

Section 2 measures recognition of selected structuraland grammatical points in standard written English.The language tested is formal, rather than conversa-tional. The topics of the sentences are of a generalacademic nature so that individuals in specific fieldsof study or from specific national or linguistic groupshave no particular advantage. When topics havea national context, they refer to United States orCanadian history, culture, art, or literature. However,knowledge of these contexts is not needed to answerthe structural or grammatical points being tested.

This section is divided into two parts. The firstpart tests an examinee’s ability to identify the correctstructure needed to complete a given sentence. Theexaminee reads incomplete sentences printed in thetest book. From the four responses provided for eachincomplete sentence, the examinee must choose theword or phrase that best completes the given sentence.Only one of the choices fits correctly into the particularsentence. The second part tests an examinee’s abilityto recognize correct grammar and to detect errors instandard written English. Here the examinee readssentences in which some words or phrases are under-lined. The examinee must identify the one underlinedword or phrase in each sentence that would not beaccepted in standard written English.

n Section 3, Reading Comprehension

Section 3 measures the ability to read and understandshort passages that are similar in topic and style tothose that students are likely to encounter in NorthAmerican colleges and universities. The examineereads a variety of short passages on academic subjectsand answers several questions about each passage.The questions test information that is stated in orimplied by the passage, as well as knowledge of someof the specific words as they are used in the passage.To avoid creating an advantage to individuals in anyone field of study, sufficient context is provided sothat no subject-specific familiarity with the subjectmatter is required to answer the questions. Questionsare asked about factual information presented in thepassages, and examinees may also be asked to makeinferences or recognize analogies. In all cases, thequestions can be answered by reading and under-standing the passages.

Development ofTOEFL Test Questions

Material for the TOEFL test is prepared by lan-guage specialists who are trained in writing questionsfor the test before they undertake actual item-writingassignments. Additional material is prepared byETS test development specialists. The members ofthe TOEFL Committee of Examiners establish overallguidelines for the test content and specifications. Allitem specifications, questions, and final test forms arereviewed internally at ETS for cultural and racial biasand content appropriateness, according to establishedETS procedures.

These reviews ensure that each final form of thetest is free of any language, symbols, references, orcontent that might be considered potentially offensiveor inappropriate for subgroups of the TOEFL testpopulation, or that might serve to perpetuate negativestereotypes.

All questions are pretested on representativegroups of international students who are not nativespeakers of English. Only after the results of thepretest questions have been analyzed for statisticaland content appropriateness are questions selectedfor the final test forms.

Following the administration of each new formof the test, a statistical analysis of the responses toquestions is conducted. On rare occasions, whena question does not function as expected, it will bereviewed again by test specialists. After this review,the question may be deleted from the final scoringof the test. The statistical analyses also providecontinuous monitoring of the level of difficulty ofthe test, the reliability of the entire test and of eachsection, intercorrelations among the sections, andthe adequacy of the time allowed for each section.(See “Statistical Characteristics of the Test,” page 29.)

13

TOEFL TESTING PROGRAMS ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

The TOEFL test is administered internationally onregularly scheduled test dates through the Fridayand Saturday testing programs. It is also administeredat local institutions around the world throughthe Institutional Testing Program (ITP). The ITPprogram does not provide official TOEFL scorereports; scores are for use by the administeringinstitution only.

Friday and SaturdayTesting ProgramsThe official TOEFL test is given at centers aroundthe world one day each month – five Fridays andseven Saturdays.

The TOEFL office diligently attempts to make thetest available to all individuals who require TOEFLscores. In 1996-97, more than 1,275 centers locatedin 180 countries and areas were established for theSaturday testing program to accommodate the morethan 703,000 persons registered to take the test; 350centers in more than 60 countries and areas wereestablished for the more than 248,000 personsregistered to take TOEFL under the Friday program.

Registration and administration procedures areidentical for the Friday and Saturday programs. Thetest itself is also identical in terms of format andcontent. Score reports for administrations under bothprograms provide the same data. More informationabout these testing programs can be found in theBulletin of Information for TOEFL, TWE, and TSE.(See page 47.)

As noted above, the TOEFL program provides12 test dates a year. However, the actual number ofadministrations at any one center in a given countryor area is scheduled according to demand and theavailability of space and supervisory staff.

There are sometimes local scheduling conflictswith national or religious holidays. Although theTOEFL office makes every effort to avoid schedulingadministrations of the test on such dates, it may beunavoidable in some cases.

Registration must be closed well in advance of eachtest date to ensure the delivery of test materials to thetest centers. Registration deadline dates are aboutseven weeks before the test dates for centers outsidethe United States and Canada and five weeks beforethe test dates for centers within these two countries.

Almost all administrations are held as scheduled.On occasion, however, shipments of test materials maybe impounded by customs officials or delayed by mailembargoes or transportation strikes. Other problems,ranging from political disturbances within countries,to power failures, to the last-minute illness of a testsupervisor, may also force postponement of a TOEFLtest administration.

If an administration must be postponed, a makeupadministration is scheduled, usually on the nextregularly scheduled test date. Occasionally it isnecessary to arrange a makeup administration onanother date.

Different forms of the test may be used at a singleadministration. Following each administration, theanswer sheets are returned to ETS for scoring; testresults are mailed to score recipients about one monthafter the answer sheets are received at ETS.

TWE Test (Test of Written English)In 1986, the TOEFL program introduced the Test ofWritten English. This direct assessment of writingproficiency was developed in response to requestsfrom many colleges, universities, and agencies thatuse TOEFL scores. The TWE test is currently(1997) a required section of the TOEFL test at fiveadministrations per year. For more informationabout the Test of Written English, see page 39.

TSE Test (Test of Spoken English)The Test of Spoken English measures the ability ofnonnative speakers of English to communicate orallyin English. It requires examinees to tape recordspoken answers to a variety of questions. The TSEtest is administered on all 12 Friday and SaturdayTOEFL test dates. For more information about theTest of Spoken English, see page 39.

Institutional Testing ProgramThe Institutional Testing Program permits approvedinstitutions throughout the world to administer theTOEFL test to their own students on dates conven-ient for them (except for regularly scheduled TOEFLadministration dates), using their own facilities andstaff. Each year a number of forms of the TOEFL testpreviously used in the Friday and Saturday testingprograms are made available for the InstitutionalTesting Program.

14

In addition to the regular TOEFL test, which isespecially appropriate for use with students at theintermediate and higher levels of English languageproficiency, ITP offers the Preliminary Test of Englishas a Foreign Language (Pre-TOEFL) for individuals atthe beginning level. Pre-TOEFL measures the samecomponents of English language skills as the TOEFLtest. However, Pre-TOEFL is less difficult and shorter.Pre-TOEFL test results are based on a restricted scalethat provides more discriminating measurement atthe lower end of the TOEFL scale.

Note: There are minor differences in the numberof questions and question types between the ITPTOEFL test and the Pre-TOEFL test.

How Institutional TOEFL Can Be Used

The Institutional Testing Program is offered primarilyto assist institutions in placing students in Englishcourses at the appropriate level of difficulty, fordetermining whether additional work in Englishis necessary before an individual can undertakeacademic studies, or as preparation for an officialFriday or Saturday TOEFL administration.

Institutional TOEFL Test Scores

Scores earned under the Institutional TestingProgram are comparable to scores earned under theworldwide Friday and Saturday testing programs.However, ITP scores are for use by the administeringinstitution only.

ETS reports test results to the administeringinstitution in roster form, listing the names andscores (section and total) of all students who took thetest at that administration. Two copies of the scorerecord for each student are provided to the administer-ing institution: a file copy for the institution and apersonal copy for the individual. Both copies indicatethat the scores were obtained at an InstitutionalTesting Program administration.

ETS does not report scores obtained under thisprogram to other institutions as it does for officialscores obtained under the Friday and Saturday testingprograms. To ensure score validity, scores obtainedunder the Institutional Testing Program shouldnot be accepted by other institutions to evaluate anindividual’s readiness to begin academic studiesin English.

15

PROCEDURES AT TEST CENTERS ○ ○ ○ ○ ○

Standard, uniform procedures are important in anytesting program, but are essential for an examinationthat is given worldwide. Therefore, the TOEFLprogram provides detailed guidelines for test centersupervisors to ensure uniform administrations.Preparing for a TOEFL/TWE or TSE Administrationis mailed to test supervisors well in advance of the testdate. This publication describes the arrangements thesupervisor must make to prepare for the test adminis-tration, including selecting testing rooms and theassociate supervisors and proctors who will be neededon the day of the test.

The Manual for Administering TOEFL, includedwith every shipment of test materials, describesappropriate seating plans, the kind of equipment thatshould be used for the Listening Comprehensionsection, identification requirements, the priorities foradmitting examinees to the testing room, and instruc-tions for distributing and collecting test materials.It also contains detailed instructions for the actualadministration of the test.

TOEFL program staff work with test center super-visors to ensure that the same practices are followed atall centers, and they conduct workshops during whichsupervisors can discuss procedures for administering thetest. TOEFL staff respond to all inquiries from supervi-sors and examinees regarding circumstances or condi-tions associated with test administrations, and theyinvestigate all complaints received about specificadministrations.

Measures to Protect Test SecurityIn administering a worldwide testing program at morethan 1,275 test centers in 180 countries, the TOEFLprogram considers the maintenance of security attesting sites to be of paramount importance. Theelimination of problems at test centers, including test-taker impersonations, is a continuing goal. To offerscore users the most valid, reliable, and secure mea-surements of English language proficiency available,the TOEFL office continuously reviews and refinesprocedures to increase the security of the test before,during, and after its administration.

Because of the importance of TOEFL test scores toexaminees and institutions, it is inevitable that someindividuals will engage in practices designed to increasetheir reported scores. The careful selection of supervi-sors, a high proctor-to-examinee ratio, and carefullydeveloped procedures for the administration of the test

(explained in the Manual for Administering TOEFL)are measures designed to prevent or discourage exam-inee attempts at impersonation, copying, theft of testmaterials, and the like, and thus to protect the integrityof the test for all examinees and score recipients.

Identification Requirements

Strict admission procedures are followed at all testcenters to prevent attempts by some examinees tohave others with greater proficiency in Englishimpersonate them at a TOEFL administration. Tobe admitted to a test center, every examinee mustpresent an official document with a recognizablephotograph and a completed photo file record with arecent photo attached. Although the passport is thebasic document that is acceptable at all test centers,other specific photobearing documents may be accept-able for individuals who may not be expected to havepassports or who are taking the test in their owncountries.

Through embassies in the United States andTOEFL representatives and supervisors in othercountries, the TOEFL office continually verifies thenames of official, secure, photobearing identificationdocuments used in each country, such as nationalidentity cards, work permits, and registration certifi-cates. In the Friday and Saturday testing programs,each admission ticket contains a statement specifyingthe documents that will be accepted at TOEFL testcenters in the country in which the examinee isregistered to take the test. This information is com-puter-printed on a red field to ensure that it will beseen. (The same information is printed on theattendance roster prepared for each center.) Followingis a sample of the statement that appears on admis-sion tickets for Venezuela.

YOUR VALID PASSPORT. CITIZENS OF VEN-EZUELA MAY USE NATIONAL IDENTITY CARDOR LETTER AS DESCRIBED IN THE BULLETIN.

Complete information about identification require-ments is included in all editions of the Bulletin ofInformation for TOEFL, TWE, and TSE.

16

Photo File Records

Every TOEFL examinee must present a completedphoto file record to the test center supervisor beforebeing admitted to the testing room. The photo filerecord contains the examinee’s name, registrationnumber, test center code, and signature, as well asa recent photo that clearly identifies the examinee(that is, the photo must look exactly like the exam-inee, with the same hairstyle, with or without a beard,and so forth). The photo file records are collected atthe test center and returned to ETS, where the photosand identifying information are electronically cap-tured and included on the examinee’s score data file.

Photo Score Reporting

As an additional procedure to help eliminate thepossibility of impersonation at test centers, theofficial score reports that are routinely sent to institu-tions designated by the test taker, and the examinee’sown copy of the score report, bear an electronicallyreproduced photo image of the examinee and his orher signature. (The score report also includes the

number of the passport or other identificationdocument used to gain admission to the testing centerand the name of the country issuing the document.)Examinees are advised in the Bulletin of Informationthat the score reports will contain these photo images.In addition to strengthening security through thisdeterrent to impersonation, the report form providesscore users with the immediate information they mayneed to resolve any issues of examinee identity. Keyfeatures of the image score reports are highlighted onpage 19.

Checking Names

To prevent examinee attempts to exchange answersheets or to grid another person’s name (for whom heor she is taking the test) on the answer sheet, supervi-sors are asked to compare names on the identificationdocument and the answer sheet and also to check thegridding of names on the answer sheet before examin-ees leave the room.

16

17

Supervision of Examinees

Supervisors and proctors are instructed to exerciseextreme vigilance during a test administration toprevent examinees from giving or receiving assistancein any way.

In addition, the Manual for Administering TOEFLadvises supervisors about assigning seats to examin-ees. To prevent copying from notes or other aids,examinees may not have anything on their desks buttheir test books, answer sheets, pencils, and erasers.They are not permitted to make notes or marks of anykind in their test books. (Warning/Dismissal Noticeforms are used to report examinees who violateprocedures. An examinee is asked to sign the notice todocument the violation and to indicate he or sheunderstands that a violation of procedures hasoccurred and that the answer sheet may not bescored.)

If a supervisor is certain that someone has given orreceived assistance, the supervisor has the authorityto dismiss the examinee from the testing room; scoresfor dismissed examinees will not be reported. If asupervisor suspects someone of cheating, the exam-inee is warned about the violation, is asked to sign aWarning/Dismissal Notice, and must move to anotherseat selected by the supervisor. A description of theincident is written on the Supervisor’s IrregularityReport, which is returned to ETS with the answersheet. Both suspected and confirmed cases of cheatingare investigated by the Test Security Office at ETS.(See “Scores of Questionable Validity,” page 23.)

Turning back to another section of the test,working on a section in advance, or continuing towork on a section after time is called are not permit-ted and are considered cheating. (To assist thesupervisor, a large number identifying the sectionbeing worked on is printed at the top of each page ofthe test book.) Supervisors are instructed to warnanyone found working on the wrong section and toask the examinee to sign a Warning/Dismissal Notice.

Preventing Access toTest Materials

To ensure that examinees have not seen the testmaterial in advance, a new form of the test is devel-oped for each Friday and Saturday administration.

To prevent the theft of test materials, procedureshave been devised for the distribution and handlingof these materials. Test books are individually sealed,then packed and sealed in plastic bags. Test books,answer sheets, and Listening Comprehension record-ings are sent to test centers in sealed boxes and areplaced in secure, locked storage that is inaccessible tounauthorized persons. Supervisors are directed tocount the test books several times — upon receipt,during the test administration, and after the test isover. No one is permitted to leave the testing roomuntil the supervisor has accounted for all test materi-als. Except for “disclosed” administrations, whenexaminees may obtain the test book (see “Test FormsAvailable to TOEFL Examinees,” page 47), supervi-sors must follow detailed directions for returning thetest materials. Materials are counted upon receipt atETS, and its Test Security Office investigates all casesof missing test materials.

18

TOEFL TEST RESULTS ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

Release of Test ResultsAbout one month after a Friday or Saturday TOEFLadministration, test results are mailed to the examin-ees and to the official score recipients they havespecified, provided that the answer sheets are receivedat ETS promptly after the administration. Test resultsfor examinees whose answer sheets are incomplete orwhose answer sheets arrive late are usually sent twoor three weeks later. All test results are mailed by thefinal deadline — 12 weeks after the test.

For the basic TOEFL test fee, each examinee isentitled to four copies of the test results: one copyis sent to the examinee, and up to three official scorereports are sent directly by ETS to the institutionswhose assigned code numbers the examinee hasmarked on the answer sheet.* The institution codedesignates the recipient college, university, or agency.A list of the most frequently used institution andagency codes is printed in the Bulletin of Information.An institution whose code number is not listed shouldgive applicants its code number before they take thetest. (See page 20 for more information.)

The most common reason that institutions do notreceive score reports following an administration isthat examinees do not properly specify the institu-tions as score report recipients by marking the correctcodes on the test answer sheet. (Examinees cannotwrite the names of recipients on the answer sheet.)An examinee who wants scores sent to an institutionwhose code number was not marked on the answersheet must submit a Score Report Request Formnaming the institution that is to receive the scores.There is a fee for this service.

Test Score Data RetentionLanguage proficiency can change considerably in arelatively short period. Therefore, the TOEFL officewill not report scores that are more than two yearsold. Individually identifiable TOEFL scores areretained on the TOEFL database for only two yearsfrom the date of the test. Individuals who took theTOEFL test more than two years ago must take itagain if they want scores sent to an institution.*After two years, all information that could be used toidentify an individual is removed from the database.Score data and other information that may be used forresearch or statistical purposes do not include indi-vidual examinee identification information and areretained indefinitely.

Image Score ReportsThe image-processing technology used to producethe photo score reports allows ETS to electronicallycapture the image from the examinee’s photograph,as well as the signature and other identifying datasubmitted by the examinee at the testing site, andto reproduce these with the examinee’s test resultsdirectly on the score reports. The computerizedelectronic transfer of photo images permits a high-quality reproduction of the original photo on the scorereport. (If a photograph is too damaged or for otherreasons cannot be accepted by the image-processingsystem, “Photo Not Available” will be printed on thescore report.)

Steps have been taken to reduce the opportunitiesfor tampering with examinee score records thatinstitutions may receive directly from applicants.However, to ensure that institutions receivevalid score records, we urge that admissionsofficers and others responsible for the admis-sions process accept only official score reportssent directly by ETS.

* An institution or agency that is sponsoring an examinee and has madeprior arrangements with the TOEFL office will also receive a copy of theexaminee’s official score report if the examinee has given permission tothe TOEFL office.

* A TOEFL score is measurement information and is subject to all therestrictions noted in this Manual. (These restrictions are also noted in theBulletin of Information.) The test score is not the property of the examinee.

19

YOUR SIGNATURE

NAME OF COUNTRY ISSUING PASSPORT OR IDENTIFICATION

NUMBER ON IDENTI-FICATION DOCUMENT

®

TEST OF ENGLISH AS A FOREIGN LANGUAGE

Test of English as a Foreign LanguageP.O. Box 6151 • Princeton, NJ 08541-6151 • USA

TEST OF WRITTEN ENGLISH

TOEFL SCALED SCORES

NAME (Family or Surname, Given, Middle)

EXAMINEE'S ADDRESS:

SECTION 1 SECTION 2 SECTION 3 TOTAL SCORE

TWE SCORE

REGISTRATIONNUMBER

Month YearTEST DATE

CENTERNUMBER

INTERPRETIVEINFORMATION

The face of this document has a multicolored background — not a white background.

NATIVE LANGUAGE

NATIVE COUNTRY

DEPARTMENT

Month/Day/YearDATE OF BIRTH SEX

INST.CODE

DEPT.CODE

DEGREE

REASONFOR

TAKINGTOEFL

TOEFLTAKEN

BEFORE

Facsimile reduced. Actual size of entire form, 81/2( x 11(; score report section 81/2( x 35/8(.

1

5

42 3

Official Score Reports from ETSTOEFL score reports give the score for each of thethree sections of the test and the total score. Examin-ees who take the TOEFL test during an administra-tion at which the Test of Written English is given alsoreceive a TWE score printed in a separate field on theTOEFL score report. See page 20 for informationabout the score report codes.

Features of the Image Reports:

a The blue background color quickly identifies thereport as being an official copy sent from ETS.

s The examinee’s name and scores are printed inred fields.

d Reverse type is used for printing the name andscores.

f The examinee’s photo is taken from the photo filerecord given to the test center supervisor on theday of the test and reproduced on the scorereport.

g The examinee’s signature and ID number andthe name of the country issuing identificationare reproduced from the photo file record.

h The word “copy” appears in the backgroundcolor of score reports that are photocopied usingeither a black or color image copier.

Score reports are valid only if received directly fromEducational Testing Service. TOEFL test scores areconfidential and should not be released by therecipient without written permission from the ex-aminee. All staff with access to score records shouldbe advised of their confidential nature.

If you have any reason to believe that someonehas tampered with a score report or would liketo verify test scores, please call the following toll-free number between 8:30 AM and 4:30 PM NewYork time.

800-257-9547TOEFL/TSE Services will verify the accuracy of

the scores.

6

20

Information Printed on theOfficial Score ReportIn addition to test scores, native country, nativelanguage, and birth date, the score report includesother pertinent data about the examinee and informa-tion about the test.

INSTITUTION CODE. The institution code designates the recipient college,university, or agency. A list of the most frequently used institution and agencycodes is printed in the Bulletin of Information. An institution whose code numberis not listed should give applicants its code number before they take the test.(This information should be included in application materials prepared forinternational students.)

Note: An institution that does not know its TOEFL code number or wishes toobtain one should call 609-771-7975 or write to ETS Code Control, P.O. Box6666, Princeton, NJ 08541-6666, USA.

DEPARTMENT CODE. The department code number identifies the profes-sional school, division, department, or field of study in which the graduateapplicant plans to enroll. The department code list shown below is also includedin the Bulletin of Information. The department code for all business schools is(02), for law schools (03), and for unlisted departments (99).

Fields of Graduate Study Other Than Business or LawBIOLOGICAL SCIENCES31 Agriculture32 Anatomy05 Audiology33 Bacteriology34 Biochemistry35 Biology45 Biomedical Sciences36 Biophysics37 Botany38 Dentistry39 Entomology46 Environmental Science40 Forestry06 Genetics41 Home Economics25 Hospital and Health Services

Administration42 Medicine07 Microbiology74 Molecular and Cellular Biology43 Nursing77 Nutrition44 Occupational Therapy56 Pathology47 Pharmacy48 Physical Therapy49 Physiology55 Speech-Language Pathology51 Veterinary Medicine52 Zoology30 Other biological sciences

PHYSICAL SCIENCES54 Applied Mathematics61 Astronomy62 Chemistry78 Computer Sciences63 Engineering, Aeronautical64 Engineering, Chemical65 Engineering, Civil66 Engineering, Electrical67 Engineering, Industrial68 Engineering, Mechanical69 Engineering, other71 Geology72 Mathematics73 Metallurgy75 Oceanography76 Physics59 Statistics60 Other physical sciences

Use 99 for any departmentnot listed.

HUMANITIES11 Archaeology12 Architecture26 Art History13 Classical Languages28 Comparative Literature53 Dramatic Arts14 English29 Far Eastern Languages and Literature15 Fine Arts, Art, Design16 French17 German04 Linguistics19 Music57 Near Eastern Languages and Literature20 Philosophy21 Religious Studies or Religion22 Russian/Slavic Studies23 Spanish24 Speech10 Other foreign languages98 Other humanities

SOCIAL SCIENCES27 American Studies81 Anthropology82 Business and Commerce83 Communications84 Economics85 Education (including M.A. in Teaching)01 Educational Administration70 Geography92 Government86 History87 Industrial Relations and Personnel88 International Relations18 Journalism90 Library Science91 Physical Education97 Planning (City, Community,

Regional, Urban)92 Political Science93 Psychology, Clinical09 Psychology, Educational58 Psychology, Experimental/

Developmental79 Psychology, Social08 Psychology, other94 Public Administration50 Public Health95 Social Work96 Sociology80 Other social sciences

TOEFL SCORES: Three section scores and a total score are reported for theTOEFL test. The three sections are:

Section 1 — Listening ComprehensionSection 2 — Structure and Written ExpressionSection 3 — Reading Comprehension

TEST OF WRITTEN ENGLISH (TWE): Effective July 1995, the TWE test isadministered in August, October, December, February, and May.

Scores Explanations of TWE Scores

6.0 Demonstrates clear competence in writing on both the rhetoricaland syntactic levels, though the essay may have occasional errors.

5.55.0 Demonstrates competence in writing on both the rhetorical and

syntactic levels, though the essay will probably have occasionalerrors.

4.54.0 Demonstrates minimal competence in writing on both the

rhetorical and syntactic levels.3.53.0 Demonstrates some developing competence in writing, but the

essay remains flawed on either the rhetorical or syntactic level, orboth.

2.52.0 Suggests incompetence in writing.1.51.0 Demonstrates incompetence in writing.1NR Examinee did not write an essay.OFF Examinee did not write on the assigned topic.

INTERPRETIVE INFORMATION: The date of the most current edition of theTOEFL Test and Score Manual is printed here. (This date is printed only onthe official score report.)

TEST DATE: Because English proficiency can change considerably in arelatively short period, please note the date on which the test was taken.Scores more than two years old cannot be reported, nor can they be verified.

PLANS TO WORK FOR DEGREE:1 = Yes 2 = No 0 = Not answered

REASON FOR TAKING TOEFL:1 = To enter a college or university as an undergraduate student2 = To enter a college or university as a graduate student3 = To enter a school other than a college or university4 = To become licensed to practice a profession5 = To demonstrate proficiency in English to the company for which

the examinee works or expects to work6 = Other than above0 = Not answered

NUMBER OF TIMES TOEFL TAKEN BEFORE:1 = One 3 = Three 0 =None or not2 = Two 4 = Four or more answered

21

Facsimile reduced

YOUR SIGNATURE

NAME OF COUNTRY ISSUING PASSPORT OR IDENTIFICATION

NUMBER ON IDENTI-FICATION DOCUMENT

®

TEST OF ENGLISH AS A FOREIGN LANGUAGE

TEST OF WRITTEN ENGLISH

TOEFL SCALED SCORES

NAME (Family or Surname, Given, Middle)

EXAMINEE'S ADDRESS:

Test of English as a Foreign Language • P. O. Box 6151 • Princeton, NJ 08541-6151 • USA

SECTION 1 SECTION 2 SECTION 3 TOTAL SCORE

TWE SCORE

REGISTRATIONNUMBER

Month YearTEST DATE

CENTERNUMBER SPONSOR CODE

EXAMINEE'S ORIGINAL SCORE RECORD

NATIVE LANGUAGE

NATIVE COUNTRY

Month/Day/YearDATE OF BIRTH SEX

INST.CODE

DEPT.CODE

DEGREE

REASONFOR

TAKINGTOEFL

TOEFLTAKEN

BEFORE

4 5

2

1

3

Examinee Score RecordsExaminees receive their test results on a form titledExaminee’s Score Record. These are NOT officialTOEFL score reports and should not be acceptedby institutions.

Acceptance of Test Results Not Receivedfrom ETS

Bear in mind that examinees may attempt to alterscore records. Institution and agency officials areurged to verify all TOEFL scores supplied by examin-ees. TOEFL/TSE Services will either confirm or denythe accuracy of the scores submitted by examinees.

If there is a discrepancy between the official scoresrecorded at ETS and those submitted in any form byan examinee, the institution will be requested to sendETS a copy of the score record supplied by theexaminee. At the written request of an official of theinstitution, ETS will report the official scores, as wellas all previous scores recorded for the examineewithin the last two years. Examinees are advised ofthis policy in the Bulletin, and, in signing theircompleted registration forms, they accept theseconditions. (Also see “Test Score Data Retention”on page 18.)

How to Recognize an Unofficial ScoreReport:

a ✭✭✭Examinee’s Original Score Record✭✭✭ isprinted at the bottom of the score record.

s The Examinee’s Score Record is printed on whitepaper.

How to Recognize If a Score ReportHas Been Altered:

d The last digit of the total score should end in “0,”“3,” or “7.”

f There should be no erasures. Do the shaded areasseem lighter than others, or are any of these areasblurred?

g The typeface should be the same in all areas.

22

DOs and DON’Ts

Do verify the information on an examinee’s scorerecord by calling TOEFL/TSE Services:

800-257-9547Don’t accept scores that are more than two

years old.

Don’t accept score reports from another institu-tion that were obtained under the TOEFLInstitutional Testing Program.

Don’t accept photocopies of score reports.

Additional Score ReportsIndividuals who have taken the TOEFL test atscheduled Friday or Saturday test administrationsmay request that official score reports be sent toadditional institutions at any time up to two yearsafter the date on which they took the test.

There are two score reporting services: (1) regularand (2) rush reporting. The regular service mailsadditional score reports within two weeks afterreceipt of an examinee’s Score Report Request Form.The rush reporting service mails score reports toinstitutions within four working days after a requestform has been received. There is an additional fee forthe rush service.

Confidentiality of TOEFL ScoresInformation retained in TOEFL test files about anexaminee’s native country, native language, and theinstitutions to which the test scores have been sent,as well as the actual scores, is the same as the infor-mation printed on the examinee’s score record andon the official score reports. An official score reportwill be sent only at the written consent of the exam-inee to those institutions or agencies designated onthe answer sheet by the examinee on the day of thetest, on a Score Report Request Form submitted ata later date, or otherwise specifically authorized bythe examinee.*

* See footnote on page 18.

To ensure the authenticity of scores, the TOEFLprogram office urges that institutions accept onlyofficial copies of TOEFL scores received directlyfrom ETS.

Score users are responsible for maintaining theconfidentiality of an individual’s score information.Scores are not to be released by the institutionalrecipient without the explicit permission of theexaminee. Dissemination of score records should bekept to a minimum, and all staff with access to themshould be informed of their confidential nature.

The TOEFL program recognizes the right ofinstitutions as well as individuals to privacy withregard to information supplied by and about themthat is stored in data or research files held by ETSand the concomitant responsibility to safeguardinformation in its files from unauthorized disclosure.As a consequence, information about an institution(identified by name) will be released only in a mannerconsistent with a prior agreement, or with the explicitconsent of the institution.

Calculation of TOEFL ScoresThe raw scores for the three sections of the TOEFLtest are the number of questions answered correctly.No penalty points are subtracted for wrong answers.Although each new form of the test is constructed tomatch previous forms in terms of content and diffi-culty, the level of difficulty may vary slightly from oneform to another. Raw scores from each new TOEFLtest are statistically adjusted, or equated, to accountfor relatively minor differences in difficulty acrossforms, thereby allowing scores from different formsof the test to be used interchangeably.

At the time of the first administration of the three-section TOEFL test (1976), the scale for reporting thetotal score was linked to the scale that was then in usefor the original five-section test. Since April 1996 thescale has been maintained by linking current tests tothe scale of the July 1995 initial revised TOEFL test.

The three separate sections are scaled so the meanscaled score for each section equals one-tenth of thetotal scaled score mean (the standard deviations of thescaled scores for the three sections are equal) and thetotal score equals ten-thirds times the sum of the threesection scaled scores.

23

Example:

Section 1 Section 2 Section 3 Sum46 + 54 + 50 = 150

(150 x 10) ÷ 3 = 500

TOEFL scores for Sections 1 and 2 are reported ona scale that can range from 20 to 68. Section 3 scoresrange from 20 to 67. TOEFL total scores are reportedon a scale that can range from 200 to 677.

Scores for each new test form are converted to thesame scale by a statistical equating procedure knownas item response theory (IRT) true score equating,which determines equivalent scaled scores for personsof equal ability regardless of the difficulty level of theparticular form of the test and the average ability levelof the group taking the test.*

The reported scores are not based on either thenumber or the percentage of questions answeredcorrectly. Nor are they related to the distributionof scores on any other test, such as the SAT or theGRE tests.

Actual ranges of observed scores for the periodfrom July 1995 through June 1996 are shown inTable 1. Note that for the section and total scores,all minimum observed section and total scores arehigher than the lowest possible scores.

Hand-Scoring ServiceExaminees are responsible for properly completingtheir answer sheets to ensure accurate scoring. Theyare instructed to use a medium-soft black lead pencil,to mark only one answer to each question, to fill inthe answer space completely so the letter inside thespace cannot be seen, and to erase all extra marksthoroughly. Failure to follow any of these instructionsmay result in the reporting of an inaccurate score.

Examinees who question whether their reportedscores are accurate may request that their answersheets be hand scored. There is a fee for this service.A request for hand scoring must be received withinsix months of the test date; later requests cannotbe honored.

The TOEFL office has established the followinghand-scoring procedures: the answer sheet to be handscored is first confirmed as being the one completedby the person requesting the service; the answer sheetis then hand scored twice by trained ETS staffworking independently. If there is a discrepancybetween the hand-scored and computer-scored results,the hand-scored results, which may be higher or lowerthan those originally reported, will be reported to allrecipients of the earlier scores, and the hand-scoringfee will be refunded to the examinee. The results ofthe hand scoring are available about three weeks afterreceipt of the examinee’s request. Experience hasshown that very few score changes result from hand-scoring requests.

Scores of Questionable ValidityImproved scores over time are to be expected if aperson is studying English; they may not indicateirregularities. However, institutions and other TOEFLscore recipients that note inconsistencies between testscores and English performance, especially in caseswhere there is reason to suspect an inconsistencybetween a high TOEFL score and relatively weakEnglish proficiency, are encouraged to refer to theofficial photo score report for the possibility ofimpersonation. Institutions should notify the TOEFLoffice if they find any evidence of impersonation. ETSreports TOEFL scores for a period of two years afterthe date the test was administered.

Table 1. Minimum and Maximum ObservedSection and Total Scores, July 1995 - June 1996

Section Min. Max.

1. Listening Comprehension 25 682. Structure and Written Expression 21 683. Reading Comprehension 22 67

Total Score 263 677

* See Cook and Eignor (1991) for further information about IRT truescore equating.

This method of scaling results in rounded scoresfor which the last digit can take on only three values:zero, three, or seven.

24

Irregularities uncovered by institutions andreported to ETS, as well as those brought to theattention of the TOEFL office by examinees orsupervisors who believe that misconduct may havetaken place, are investigated.

Misconduct irregularities are reviewed, statisticalanalyses are conducted, and scores may be canceledby ETS. For other irregularities, the ETS Test Secu-rity Office assembles relevant documents, such asprevious score reports, registration forms, and answersheets. When handwriting differences or evidenceof possible copying or exchange of answer sheets isfound, the case is referred to the ETS Board ofReview, a group of senior professional staff members.Based on its independent examination of the evi-dence, the Board of Review directs appropriate action.

ETS policy and procedures are designed to providereasonable assurance of fairness to examinees in boththe identification of suspect scores and the weighingof information leading to possible score cancellation.These procedures are intended to protect both scoreusers and examinees from inequities that could resultfrom decisions based on fraudulent scores and tomaintain the integrity of the test.

Examinees with DisabilitiesNonstandard testing arrangements may include specialeditions of the test, the use of a reader and/or amanuen-sis, a separate testing room, and extended time and/orrest breaks during the test administration.

Nonstandard administrations are given on regularlyscheduled test dates, and security procedures are thesame as those followed for standard administrations.

The TOEFL office advises institutions that the testmay not provide a valid measure of the examinee’sproficiency, even though the conditions were designedto minimize any adverse effects of the examinee’sdisability upon test performance. The TOEFL officecontinues to recommend that alternative methods ofevaluating English proficiency be used for individualswho cannot take the test under standard conditions.Criteria such as past academic record (especially ifEnglish has been the language of instruction), recom-mendations from language teachers or others familiarwith the applicant’s English proficiency, and/or apersonal interview or evaluation are suggested in lieuof TOEFL scores. Because the individual circum-stances of nonstandard administrations vary sowidely and the number of examinees tested undernonstandard conditions is still quite small, theTOEFL program cannot provide normative data forinterpreting scores obtained in such administrations.

A statement that the scores were obtained undernonstandard conditions is printed on the official scorereport (and on the Examinee’s Score Record) of anexaminee for whom special arrangements were made.Each score recipient is also sent an explanatory noticeemphasizing that there are no normative data forscores obtained under nonstandard testing conditionsand, therefore, that such scores should be used withinthese parameters.

25

USE OF TOEFL TEST SCORES ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

The TOEFL test is a measure of general Englishproficiency. It is not a test of academic aptitudeor of subject matter competence, nor is it a directtest of English speaking or writing ability.TOEFL test scores can assist in determining whetheran applicant has attained sufficient proficiency inEnglish to study at a college or university. However,even though an applicant may achieve a high TOEFLscore, the student who is not academically preparedmay not easily succeed in a given program of study.Therefore, determination of academic admissibilityof nonnative English speakers is dependent uponnumerous additional factors, such as previous aca-demic record, other institution(s) attended, level andfield of study, and motivation.

If a nonnative English speaker meets academicrequirements, official TOEFL test scores may be usedin making the following kinds of decisions:

n The applicant may begin academic work with norestrictions.

n The applicant may begin academic work with somerestrictions on academic load and in combinationwith concurrent work in English language classes.(This implies that the institution can provide theappropriate English courses to complement theapplicant’s part-time academic schedule.)

n The applicant is declared eligible to begin anacademic program within a stipulated period oftime but is assigned to a full-time program ofEnglish study. (Normally, such a decision is madewhen an institution has its own intensive English-as-a-second-language program.)

n The applicant’s official status will not be deter-mined until he or she reaches a satisfactory level ofEnglish proficiency. (Such a decision will requirethat the applicant pursue full-time English train-ing, at the same institution or elsewhere.)

All of the above decisions require the institutionto judge whether the applicant has sufficientcommand of English to meet the demands of aregular or modified program of study. Suchdecisions should never be based on TOEFLscores alone; they should be based on all rel-evant information available.

Who Should Take the TOEFL Test?All international applicants who are nonnativespeakers of English should provide evidence of theirlevel of English proficiency prior to beginning aca-demic work at an institution where English is thelanguage of instruction. TOEFL scores are frequentlyrequired for the following categories of applicants:

n Individuals from countries in which English is oneof the official languages, but not necessarily thefirst language of the majority of the population orthe language of instruction at all levels of school-ing. Such countries may include, but are notlimited to, the British Commonwealth countriesand US territories and possessions.

n Persons from countries where English is not thenative language, even though there may be schoolsor universities in which English is the language ofinstruction.

Many institutions report that they frequently donot require TOEFL test scores of certain kinds ofinternational applicants. These include:

n Nonnative speakers who hold degrees or diplomasfrom postsecondary institutions in English-speaking countries (e.g., the United States, Canada,England, Ireland, Australia, New Zealand),provided they have spent a specified minimumperiod of time in successful full-time study(generally two years) with English as the languageof instruction.

n Transfer students from other institutions in theUnited States or Canada after favorable evaluation ofprevious academic course work and course load andlength of time at the previous institution.

n Nonnative speakers who have taken the TOEFLtest within the past two years and who havesuccessfully pursued academic work in an English-speaking country for a specified minimum periodof time (generally two years) with English as thelanguage of instruction.

26

Guidelines for UsingTOEFL Test ScoresAs part of its general responsibility for the tests itproduces, the TOEFL program is concerned about theuse of TOEFL test scores by recipient institutions.The program office makes every effort to ensurethat institutions use TOEFL scores properly — forexample, by providing this Manual to all institutionsthat are interested in using the scores and by regularlyadvising institutions of any program changes that mayaffect the interpretation of TOEFL test scores. TheTOEFL office encourages individual institutions torequest assistance of TOEFL professional staff relatingto the proper use of scores.

An institution that uses TOEFL test scores shouldconsider certain factors to evaluate an individual’sperformance on the test and to determine appropriatescore requirements. The following guidelines arepresented to assist institutions in arriving at reason-able decisions.

n Base the evaluation of an applicant’sreadiness to begin academic work on allavailable relevant information, not solelyon TOEFL test scores.

The TOEFL test measures an individual’s ability inseveral areas of English language proficiency. It isnot designed to provide information about scholasticaptitude, motivation, language-learning aptitude,or cultural adaptability. The eligibility of a foreignapplicant should be fully established on the basisof all relevant academic and other criteria, includingsufficient proficiency in English to undertake theacademic program at that institution.

n Do not use rigid cut-off scores to evaluate anapplicant’s performance on the TOEFL test.

Because test scores are not perfect measures of ability,the use of rigid cut-off scores should be avoided. Thestandard error of measurement should be understoodand taken into consideration in making decisionsabout an individual’s test performance or in establish-ing appropriate critical score ranges for theinstitution’s academic demands (see “Reliabilities andthe Standard Error of Measurement,” page 29).

* See page 39 for information about the Test of Spoken English andoral proficiency.

n Consider TOEFL section scores as well astotal scores.

The total score on the multiple-choice TOEFL testis based on the scores of the three sections of the test.Although a number of applicants may achieve thesame total score, they may have different section scoreprofiles, which could significantly affect subsequentacademic performance. For example, an applicant witha low score on the Listening Comprehension sectionbut relatively high scores on the other sections mighthave greater initial difficulty in lecture courses.*This information could be used in advising andplacing applicants.

If an applicant’s score on the Structure and WrittenExpression section is considerably lower than thescores on the other sections or if the applicant’s scoreon the TWE test is low, it may be that the individualshould take a reduced academic load or be placed ina course designed to improve composition skills andknowledge of English grammar. An applicant whosescore on the Reading Comprehension section is muchlower than the scores on the other two sections mightbe advised to take a reduced academic load or topostpone enrollment in courses that involve a signifi-cant amount of reading.*

n Consider the kinds and levels of Englishproficiency required in different fields andlevels of study and the resources available atthe institution for improving the Englishlanguage skills of nonnative speakers.

An applicant’s field of study can affect the kind andlevel of language proficiency that are appropriate.Students pursuing studies in fields requiring highverbal ability (such as journalism) will need a greatercommand of English, particularly structure andwritten expression and writing, than will those infields that are not so dependent upon reading andwriting abilities. Many institutions require a higherrange of TOEFL test scores for graduate applicantsthan for undergraduates.

Institutions offering courses in English for nonna-tive speakers of English can modify academic courseloads to allow for additional concurrent languagetraining, and thus may be able to consider applicantswith a lower range of scores than can institutions thatdo not offer additional language training.

* See page 39 for information about TSE.

27

* Chase and Stallings, 1966; Heil and Aleamoni, 1974; Homburg, 1979;Hwang and Dizney, 1970; Odunze, 1980; Schrader and Pitcher, 1970;Sharon, 1972.

** A separate publication, “Guidelines for TOEFL Institutional ValidityStudies,” provides information to assist institutions in the planning oflocal validity studies. This publication is available without charge fromthe TOEFL program office upon request.

n Consider TOEFL test scores to helpinterpret an applicant’s performance onother standardized tests.

International applicants are frequently required totake standardized admission tests in addition to theTOEFL test. In such cases, TOEFL scores may proveuseful in interpreting the scores obtained on the othertests. For example, if an applicant’s TOEFL scoresare low and the scores on another test are also low(particularly one that is primarily a measure ofaptitude or achievement in verbal areas), one canlegitimately infer that the applicant’s performanceon the other test was impaired because of deficienciesin English. On the other hand, application recordsof students with high verbal aptitude scores but lowTOEFL scores should be reviewed carefully. Thescores may not be valid.

Interpreting the relationship between the TOEFLtest and aptitude and achievement tests in verbalareas can be complex. Few of even the most qualifiedforeign applicants approach native proficiency inEnglish. Factors such as cultural differences ineducational programs may also affect performanceon tests of verbal ability.

The TOEFL program has published four researchreports that can assist in evaluating the effect oflanguage proficiency on an applicant’s performanceon specific standardized tests.

The Performance of Nonnative Speakers of English onTOEFL and Verbal Aptitude Tests (Angelis, Swinton,and Cowell, 1979) gives comparative data aboutforeign student performance on TOEFL and eitherthe GRE verbal or the SAT verbal and the Test ofStandard Written English (TSWE). It providesinterpretive information about how combined testresults might best be evaluated by institutions that areconsidering foreign students. The Relationship BetweenScores on the Graduate Management Admission Testand the Test of English as a Foreign Language (Powers,1980) provides a similar comparison of performanceon the GMAT and TOEFL tests. Finally, LanguageProficiency as a Moderator Variable in Testing Aca-demic Aptitude (Alderman, 1981) and GMAT andGRE Aptitude Test Performance in Relation to PrimaryLanguage and Scores on TOEFL (Wilson, 1982)contain information supplementing that providedin the other two studies. (See “Validity,” page 34.)

n Do not use TOEFL test scores to predictacademic performance.

The TOEFL test is designed to be a measure ofEnglish language proficiency, not of academic apti-tude. Although there may be some unintended overlapbetween language proficiency and academic aptitude,other tests have been designed to measure academicaptitude more precisely and are available for thatpurpose. Use of TOEFL scores to predict academicperformance is inappropriate. Numerous predictivevalidity studies,* using grade-point averages ascriteria, have been conducted in the past. Thesestudies have shown that correlations between TOEFLtest scores and grade-point averages are often too lowto be of any practical significance. Moreover, lowcorrelations are to be expected when TOEFL scoresare used properly. If an institution admits thoseinternational applicants who have demonstrated ahigh level of language competence, one would expectthat English proficiency would no longer be highlycorrelated with academic success.

The English proficiency of an internationalapplicant is not as stable a characteristic as verbal ormathematical aptitude. Proficiency in a language issubject to change over relatively short periods of time.If considerable time has passed between the date onwhich an applicant took the TOEFL test and the dateon which he or she actually begins academic studies,there may be a greater impact on academic perfor-mance due to language loss than had been anticipated.On the other hand, a student who might be disadvan-taged because of language problems during the firstterm of study might not be disadvantaged in subse-quent terms.

n Assemble information about the validityof TOEFL test score requirements at theinstitution.

The TOEFL program strongly encourages users todesign and carry out institutional validity studies.**Because it