Comparison of the TOEFL Internet-based Test (iBT) and Paper-based Test (PBT)
The TOEFL internet-based test (iBT) and paper-based test (PBT) differ substantially in terms of availability and administration. The iBT was developed in the late 1990s and introduced to the public in 2005, and in the years since it has almost completely replaced the PBT. More than 97% of students now take the TOEFL iBT, and the PBT continues to exist only because internet testing is not yet universally available. Most countries around the world offer the iBT exclusively, and the convenience of internet testing means that it is much easier to find a test date for the iBT than for the PBT. The iBT can be taken on almost any Friday or Saturday of the year, while the few locations that offer the PBT administer the exam on a monthly basis at most. The lesser availability of the PBT is also explained by the fact that paper tests are more expensive to proctor, transport, and assess.
The iBT and the PBT also have significant differences of content. The iBT evaluates spoken English with a dedicated section, while the PBT does not assess speaking skills at all. The PBT's Structure and Written Expression section, in which test-takers must complete or correct sentences, has no counterpart on the iBT. The iBT writing section, 2 tasks in 60 minutes, is double the length of the PBT's Test of Written English (TWE), and with two distinct writing tasks rather than a single essay, the iBT measures a greater range of writing skills. In general, it can be said that the iBT requires test-takers to demonstrate the ability to synthesize English-language skills, and the PBT is a much more limited assessment (since the former was designed to improve on the latter, this is to be expected). The iBT takes a greater length of time to complete than the PBT (the iBT can last from three hours and 20 minutes to four hours and 10 minutes, compared to two hours and 20-30 minutes for the PBT). The variability in test timing is partially attributable to the inclusion of experimental questions, which can differ in number across TOEFL administrations, and the larger range of iBT test times suggests more liberal use of these experimental questions on the iBT.
The scoring system for the PBT involves conversion of correct answers for each section (raw scores) into scaled scores, which are then used to calculate total scores through a simple mathematical formula. PBT sectional scores are reported from 31-68 (Listening Comprehension and Structure and Written Expression), 31-67 (Reading Comprehension), and 1-6 (TWE). The range for the PBT total score is 310 to 677. Scoring for the iBT is far less complex, with four sectional scores of 0-30 each (reading, writing, speaking, and listening) and a total score that is the sum of all four sectional scores (0-120). Educational Testing Service (ETS) has published data on the iBT and PBT that establish approximate score concordances. A total iBT score of 84 is in the 51st percentile of all test-takers, and total iBT scores of 100 and 116 represent the 81st and 99th percentiles respectively. The roughly equivalent total PBT scores are 520 (47th percentile), 600 (87th percentile), and 660 (99th percentile).
Some universities in the English-speaking world will accept either the TOEFL iBT or PBT as proof of the requisite language skills, although departmental policies may vary. The Graduate School of Education at Harvard University has no official preference for either version. Applicants must receive either sectional iBT scores of 26 each and a total iBT score of 104 or sectional PBT scores of 60 (Reading Comprehension), 63 (Structure and Written Expression), 60 (Listening Comprehension), and 5.0 (TWE) and a total PBT score of 613. Other schools have more restrictive PBT policies. Princeton University's Department of Computer Science will only accept the PBT if the iBT is not offered in an applicant's home country. This department has no official TOEFL minimum standard, but the average TOEFL iBT total score for accepted students is 108. The Division of the Physical Sciences at the University of Chicago is an example of an institution that recognizes the iBT only. International undergraduate or graduate students should review the websites of the programs they are considering for details on their TOEFL preferences and policies. Test-takers may safely assume that the iBT will be accepted at almost any university in the English-speaking world. The PBT is still widely used in university admissions, but its acceptance is not universal.
The differences in test design, content, and structure severely limit the usefulness of the PBT as a practice tool for the iBT. As discussed above, the PBT contains a great deal of material that has no equivalent on the iBT, and vice versa. There are some similarities in the essay writing and listening portions of both tests, but students would be well advised to focus their preparation efforts on iBT-based diagnostic tests, practice tests, and exercises.