During your GMAT CAT, the computer-adaptive testing (CAT) system will continually adapt to your ability level, using your responses to build a customized exam for you as you go. This page explains how it all works, and how the scoring system accounts for the adaptive nature of the test.
[Related pages]
The GMAT CAT Adapts to Your Ability Level
During each of the two multiple-choice sections (Quantitative and Verbal) of the GMAT, the computerized testing system will select your questions from a large pool based on your responses to earlier questions of the same type — for example, Reading Comprehension, Critical Reasoning, or Data Interpretation. Although each GMAT test-taker is presented a unique combination of questions for the two multiple-choice sections, the topics and skill areas covered are similar for every test-taker. For each question type, the first question posed will be average in difficulty level. If you respond correctly to a question, the next question of that type will be more difficult; conversely, if you respond incorrectly, the next question of that type will be easier. As you proceed through a test section you'll encounter fewer and fewer questions of that type that are especially easy or difficult for you. So the CAT system "homes in" on your level of ability based on your input; the end result is that the GMAT CAT can measure your abilities with fewer questions than a non-adaptive test can. NOTE: No adaptive feature is incorporated into the two GMAT essay sections; for each of these test sections, the CAT randomly selects one prompt from a large pool.
The Computer-Adaptive System and GMAT Scoring
How does the GMAT scoring system account for the fact that each test-taker is presented a unique combination of questions? Well, every question in the CAT's database of Quantitative and Verbal questions is rated by difficulty level and categorized by type — the primary skill or ability the question is designed to measure. Your Quantitative and Verbal scores are each based on three factors:- how many
questions you answer correctly - the difficulty level of the questions you answer correctly
- the range of question types and skills tested among the questions you answer correctly.
Thus if test-taker X responds correctly to fewer questions than test-taker Y does, X might nevertheless score higher than Y — by responding correctly to more difficult (and generally more time-consuming) questions. By the same token, if the overall difficulty level of the questions to which X and Y respond correctly is the same, and if the number of correct responses is the same for both, then the test-taker whose correct responses are more evenly distributed among the different question types will score higher.
Procedural Features Inherent to Computer-Adaptive GMAT Testing
The GMAT CAT does not allow you to skip any Verbal or Quantitative questions; you must confirm your response to each question presented before the test allows you to proceed to the next one. Nor does the CAT allow you to return to a previous question. The reason for these constraints has to do with computer-adaptive nature of the exam. The difficulty level of questions presented to you depends on how you responded to earlier questions. If the CAT were to permit you to skip or return to questions the process by which the CAT builds your test and determines your score would be undermined.
 See also, these
THREE
related pages:
|
Find at this site:
