GMAT Scoring System
The verbal and quantitative sections of the GMAT are adaptive on a question-by-question basis. The questions on these GMAT sections are classified by the probability of students at specific score levels answering those questions correctly (the test writers calculate that a student who scores x will have a y% chance of getting a given question right). Answering a certain question correctly produces a subsequent question that is more difficult, while incorrect answers result in easier subsequent questions. The scaled score on these sections is calculated from the difficulty level of both correct and incorrect answers. Test-takers are rewarded more for correct answers on difficult questions, and are conversely penalized more for incorrect answers on easier questions. Because of this question-by-question adaptation, students who provide incorrect answers to several consecutive questions will receive lower scores than test-takers with the same number of non-consecutive mistakes. Furthermore, leaving a question unanswered results in a larger score penalty than an incorrect answer.
The integrated reasoning section does not use computer adaptation. Scores for this section are derived almost entirely from the number of correct answers. Only two outcomes are possible for two and three-part problems: either all parts of the question are correct, or no credit is received. GMAC will not reveal how the 12 integrated reasoning questions produce a score of 1 to 8, making it impossible to determine the exact point value of each question. However, the integrated reasoning section includes experimental questions that do not count toward the final score. GMAC also refuses to disclose the exact number of experimental questions, but they can number between two and four of the total. External evidence suggests that there is a small amount of score scaling to the difficulty of the questions answered correctly, but that each question nonetheless has roughly the same value.
Analytical writing scores are averaged from the assessments of a qualified human scorer and a computer program, reported in half-point intervals between 0 and 6. A score of 0 is reserved for essays that are either unrelated to the topic or not completed. Before implementing automated essay scoring, GMAC studied the issue, and they claim that the assessments of the computer program were very similar to those of the human graders in most areas. In the event that the scores given by the computer program differ by more than one point from the scores of the human grader, a second person is brought in to grade the essay, and the analytical writing scores in these cases will be the average of two human graders.
Some standardized tests penalize students for guessing, but the GMAT is not among them. In fact, not only are students not penalized for guessing, on the multiple-choice sections of the test they are penalized for not guessing. Unanswered questions adversely affect scores more than incorrect answers. This means that test-takers will benefit from answering all questions, even if these are only random guesses entered haphazardly as time is running out. On the verbal and quantitative sections of the GMAT, students should make every effort to avoid consecutive incorrect answers, because their scores will suffer more than from non-consecutive errors. This is best accomplished with disciplined time management. Students should not obsess over difficult questions and should not devote an inordinate amount of time to an individual problem. Due to the complexities of the adaptive algorithm, even test-takers who receive a perfect score have not necessarily answered every single question correctly. For the analytical writing section, it is important to stick to the topic at hand. As noted above, even the lowest quality essay will receive a score above zero if it is relevant to the essay prompt.
Most available evidence suggests that business schools view total scores and quantitative section scores as highly important to their admissions decisions. The value of the verbal, integrated reasoning, and analytical writing scores is more difficult to assess. Because high verbal scores are less common than high quantitative scores, admissions committees are likely to be impressed by strong verbal skills, and this may tip the scales in a candidate's favor. The integrated reasoning section is still rather new, and its credibility is still developing. The analytical writing score is clearly not as important as the verbal and quantitative scores. However, low integrated reasoning and/or analytical writing scores can harm an applicant, and students who can perform well on these GMAT sections will derive at least some benefit.