Microsoft s Testing Formats

Microsoft's Testing Formats

Currently, Microsoft uses four different testing formats:

• Case study

• Fixed length

• Adaptive

• Short form

As mentioned earlier, the case study approach is used with Microsoft's design exams, such as the one covered in this book. These exams consist of a set of case studies that you must analyze to answer questions related to the case studies. Such exams include one or more case studies (tabbed topic areas), each of which is followed by from four to ten questions. The question types for design exams and Core Four Windows 2000 exams are multiple-choice, build-list-and-reorder, create-a-tree, drag-and-connect, and select-and-place . Depending on the test topic, some exams are completely case based, whereas others are not.

Other Microsoft exams employ advanced testing capabilities that might not be immediately apparent. Although the questions that appear are primarily multiple-choice, the logic that drives them is more complex than older Microsoft tests, which use a fixed sequence of questions, called a fixed-length test . Some questions employ a sophisticated user interface, which Microsoft calls a simulation , to test your knowledge of the software and systems under consideration in a more or less "live" environment that behaves just like the original. The Exam Innovations link at www.microsoft.com/traincert/mcpexams/faq/innovations.asp includes a downloadable practice simulation.

For some exams, Microsoft has turned to a well-known technique, called adaptive testing , to establish a test-taker's level of knowledge and product competence. Adaptive exams look the same as fixed-length exams, but they discover the level of difficulty at which an individual test-taker can correctly answer questions. Test-takers with differing levels of knowledge or ability therefore see different sets of questions; individuals with high levels of knowledge or ability are presented with a smaller set of more difficult questions, whereas individuals with lower levels of knowledge are presented with a larger set of easier questions. Two individuals might answer the same percentage of questions correctly, but the test-taker with a higher knowledge or ability level will score higher because her questions are worth more.

Also, the lower-level test-taker will probably answer more questions than his more knowledgeable colleague. This explains why adaptive tests use ranges of values to define the number of questions and the amount of time it takes to complete the test.

Adaptive tests work by evaluating the test-taker's most recent answer. A correct answer leads to a more difficult question (and the test software's estimate of the test-taker's knowledge and ability level is raised). An incorrect answer leads to a less difficult question (and the test software's estimate of the test-taker's knowledge and ability level is lowered ). This process continues until the test targets the test-taker's true ability level. The exam ends when the test-taker's level of accuracy meets a statistically acceptable value (in other words, when her performance demonstrates an acceptable level of knowledge and ability) or when the maximum number of items has been presented (in which case, the test-taker is almost certain to fail).

Microsoft also introduced a short-form test for its most popular tests. This test delivers 30 questions to its takers, giving them exactly 60 minutes to complete the exam. This type of exam is similar to a fixed-length test, in that it allows readers to jump ahead or return to earlier questions and to cycle through the questions until the test is done. Microsoft does not use adaptive logic in this test but claims that statistical analysis of the question pool is such that the 30 questions delivered during a short-form exam conclusively measure a test-taker's knowledge of the subject matter in much the same way as an adaptive test. You can think of the short-form test as a kind of "greatest hits exam" (that is, the most important questions are covered) version of an adaptive exam on the same topic.

Some of the Microsoft exams can contain a combination of adaptive and fixed-length questions.

Microsoft tests can come in any one of these forms. Whatever you encounter, you must take the test in whichever form it appears; you can't choose one form over another. If anything, it pays more to prepare thoroughly for an adaptive exam than for a fixed-length or short-form exam: The penalties for answering incorrectly are built into the test itself on an adaptive exam, whereas the layout remains the same for a fixed-length or short-form test, no matter how many questions you answer incorrectly.

The biggest difference between an adaptive test and a fixed-length or short-form test is that on a fixed-length or short-form test, you can revisit questions after you've read them one or more times. On an adaptive test, you must answer the question when it's presented and have no opportunities to revisit that question thereafter.