ibt/next Generation TOEFL Score Information

Transcription

1 ibt/next Generation TOEFL Score Information ibt/next Generation TOEFL Field Test ETS conducted a field test of the new Internet-based (ibt) next generation TOEFL test to achieve three major goals: to evaluate item (question) properties, to inform decisions about score scales, and to provide a score link between the new ibt/next generation TOEFL test and the current computer-based TOEFL test (CBT). The field test also yielded percentile data about test performance, as well as test taker responses to English language competency descriptors, both of which were thought to be potentially useful to score recipients. Time and sample. From November 2003 to February 2004, 3,284 test takers participated in the field test. Participants were recruited from 30 countries in North America, Latin America, Africa, Asia, and Europe. Approximately 80 percent of the current TOEFL examinees are from these 30 countries. Test administrations. Each participant was required to take the new ibt TOEFL test (Form A) and a computer-based TOEFL test. Approximately 500 of the participants were also required to take a second form of the new test (Form B). Each participant completed an online questionnaire of demographic, selfassessment, and post-testing questions. A total of 2,720 usable responses were obtained from the 3,284 examinees who took Form A of the new test and the computer-based test. The table titled ibt/next Generation TOEFL Field Test Sample indicates the number of participants from each country, their percentage in the total study, and the percentage of individuals who took the current versions of the TOEFL test in that country between July 2002 and June Score Comparisons ETS has created score comparison tables to help score users establish recommended or required ibt/next generation TOEFL scores for their institutions. The field test provided data to compare performance on the new ibt test with performance on the CBT test. It did not compare performance between the new ibt test and the paper-based TOEFL. Paper-based and CBT score comparisons shown in the tables here were developed based upon data from 6,556 examinees who took both the paper-based and computer-based tests between November 1997 and March Score comparison tables are provided in score-to-score and range formats for total score reading listening writing total score reading listening, and writing only Although score comparisons can be useful in understanding the relationship between scores on the three versions of the TOEFL test, it is important to note that differences among the tests make it difficult to establish exact comparisons. The difference in the three versions of the test can be seen most clearly in the writing component. The new ibt Writing section is composed of two writing tasks: one independent essay and one integrated writing task. The computer-based Structure and Writing section includes multiple-choice questions and an essay. The paper-based Structure and Written Expression section consists of multiple-choice questions only, and the required essay score is reported separately from the total score. Therefore, the scores for these three sections are derived differently. Copyright 2004 by Educational Testing Service. All rights reserved. EDUCATIONAL TESTING SERVICE, ETS, the ETS logo, and TOEFL are registered trademarks of Educational Testing Service.

2 In addition, when comparing total scores, it should be kept in mind that while the new TOEFL test measures the four skills of reading, writing, listening, and speaking, neither the computer-based nor the paper-based version of the test measures speaking. Speaking Score Comparisons and ITA Score Standards There is no speaking score comparison because the TOEFL CBT test does not measure speaking. However, ETS conducted a standard setting study with international teaching assistant administrators in September 2004 to establish cut scores for International Teaching Assistants (ITAs) on the speaking portion of the new TOEFL test. The panel of 18 experts established two separate cut scores: first, a cut score for minimally acceptable speaking skills in order to have the lowest level of ITA contact with undergraduate students; and second, to establish a Speaking score that corresponds to the TSE (Test of Spoken English) score of 50. The lowest level of ITA contact with undergraduate students was set as 23 out of 30 scaled score points. The TSE score equivalent of 50 was established as 26 out of 30 scaled score points. Percentile Data Percentile rank information is based on results from 2,720 test takers who participated in the field test and who took both the new TOEFL test and TOEFL CBT. The participants English ability levels ranged from low to high to replicate the current TOEFL testing population, however, the field study group performed below the typical testing population on TOEFL CBT. Therefore, this field study percentile data should be used with great caution. Using Percentile Data Considerations for Score Users Although great care was taken in the design and administration of the next generation TOEFL field test, there are a number of important reasons why score users should view this percentile data as preliminary. Sample size: Results are based on a field sample of 2,720 participants. Motivation and performance: Participants typically are less motivated to perform well on field tests than on operational, high-stakes tests. Unfamiliarity with the test: The field test was administered to test takers who had no familiarity with the new ibt/next generation TOEFL test, which is based on current theories of communicative competence and uses integrated tasks that require examinees to combine language skills. In July 2004, ETS published a complete version of the new test on its Web site to familiarize teachers and test takers with the new test and its requirements. It is anticipated that increased familiarity will impact performance. The addition of a speaking measure: The current TOEFL test does not measure speaking ability, and in many parts of the world teachers have not emphasized the development of this skill in their classes. Although the addition of speaking on the new test will initially be challenging for some test takers, learning this skill will improve their ability to communicate effectively in an academic environment. When taking the TOEFL CBT test, the field study group performed below the typical TOEFL CBT population. Thus, the percentile data should be used with great caution, as they are not reflective of anticipated performance if the field study participants had performed similar to the overall TOEFL CBT population. Updated percentile data based on the high-stakes operational test will be published after the first testing year. Score means and standard deviations are also provided, and will be updated after the first year of testing. 2 Copyright 2004 by Educational Testing Service. All rights reserved.

3 ibt/next Generation TOEFL Field Test Sample Number of Percentage of Percentage of study study TOEFL Country participants participants volume Egypt Ethiopia Israel Kuwait United Arab Emirates France Germany Greece Italy Norway Romania Russia Spain Turkey Australia Hong Kong India Japan Korea Malaysia New Zealand Philippines Taiwan Thailand Argentina Brazil Columbia Mexico Canada United States Total 2, Copyright 2004 by Educational Testing Service. All rights reserved. 3

4 TOEFL Total Score Comparisons Score Comparison Score Comparison, cont Copyright 2004 by Educational Testing Service. All rights reserved. continued

5 TOEFL Total Score Comparisons (cont.) Score Comparison, cont Range Comparison Note: The paper-based total score does not include writing. The paper-based and computer-based total scores do not include speaking. Copyright 2004 by Educational Testing Service. All rights reserved. 5

6 Score Comparison TOEFL Score Comparisons for Reading Range Comparison Reading Reading Reading Reading Reading Reading Copyright 2004 by Educational Testing Service. All rights reserved.

7 TOEFL Score Comparisons for Listening Score Comparison Range Comparison Listening Listening Listening Listening Listening Listening Copyright 2004 by Educational Testing Service. All rights reserved. 7

8 TOEFL Score Comparisons for Writing Score Comparison New Internet- Paper-based based TOEFL Computer-based Structure and Writing Structure/Writing Written Expression Range Comparison New Internet- Paper-based based TOEFL Computer-based Structure and Writing Structure/Writing Written Expression Note: The new Internet-based TOEFL Writing section is composed of two writing tasks: one independent essay and one integrated writing task. The computer-based Structure and Writing section includes multiple-choice questions and an essay. The paper-based Structure and Written Expression section consists of multiple-choice questions only, and the required essay score is reported separately from the total score. Therefore, the scores for these three sections are derived differently. Copyright 2004 by Educational Testing Service. All rights reserved.

9 TOEFL Total Score Comparisons for and Writing Only This table presents another way of comparing scores between the new Internet-based TOEFL test and the computer-based and paper-based versions of the test. While the new TOEFL test will measure speaking ability, the computer-based and paper-based versions of the test do not. Therefore, this table compares total scores for the and Writing sections of the new test with total scores for the computer-based and paperbased versions of the test. The Internet-based scores are presented on a 0-90 scale because the Speaking section, which is worth up to 30 points, is not included in the comparison. Please note that the paper-based total score does not include Writing. Score Comparison TOEFL and Writing Only Computer-based Paper-based Score Comparison, cont. TOEFL and Writing Only Computer-based Paper-based continued Copyright 2004 by Educational Testing Service. All rights reserved. 9

10 TOEFL Total Score Comparisons for and Writing Only (cont.) Score Comparison, cont. Range Comparison TOEFL and Writing Only Computer-based Paper-based TOEFL and Writing Only Computer-based Paper-based Scale: Copyright 2004 by Educational Testing Service. All rights reserved.

11 Percentile Ranks, Means, and Standard Deviations for ibt/next Generation TOEFL Total Percentile Scale Score Rank Writing, Speaking Total Percentile Scale Score Rank Writing, Speaking continued Copyright 2004 by Educational Testing Service. All rights reserved. 11

12 Percentile Ranks, Means, and Standard Deviations for ibt/next Generation TOEFL (cont.) Total Percentile Scale Score Rank Writing, Speaking Total Percentile Scale Score Rank Writing, Speaking Total - Reading, Writing, Listening, Speaking N = 2,720 Mean = SD = Minimum Score = 6 Maximum Score = Copyright 2004 by Educational Testing Service. All rights reserved.

13 Percentile Ranks, Means, and Standard Deviations for ibt/next Generation TOEFL Reading Scale Score Percentile Rank Listening Scale Score Percentile Rank Reading N = 2,720 Mean = SD = 6.99 Minimum Score = 0 Maximum Score = 30 Listening N = 2,720 Mean = SD = 6.95 Minimum Score = 0 Maximum Score = 30 Copyright 2004 by Educational Testing Service. All rights reserved. 13

14 Percentile Ranks, Means, and Standard Deviations for ibt/next Generation TOEFL Writing Scale Score Percentile Rank Speaking Scale Score Percentile Rank Writing N = 2,720 Mean = SD = 6.67 Minimum Score = 0 Maximum Score = 30 Speaking N = 2,720 Mean = SD = 6.98 Minimum Score = 0 Maximum Score = Copyright 2004 by Educational Testing Service. All rights reserved.