"We have a continuous program for revising the exam and writing new questions. We . . . obtain comprehensive data on key professionaltasks or functions performed by planners by conducting a job analysis survey every five or so years. We use the results to link the knowledge of planners to the dimensions of their jobs, and use that information for the revision and further development of our test specifications.

"The exam is graded in two stages. We use a systematic approach to identifying and removing flawed questions from the exam. Those examinees who get high scores in the exam usually disagree when it comes to picking the answer to a flawed question. They will divide between two or even three of the possible answers as being correct. Our statistics show which questions have received such dispersed answers. If our review traces the problem to the question, we grade all answers to that one question as correct. This is done prior to scoring [the individual exams].

Your reported score is not the number of questions or the percent of questions you answered correctly. The number of questions answered correctly is converted to a scale ranging from 25 to 75. [The spread of test scores nationwide for the 1997 exam was from a low of 39 to a high of 72.] The number of correct answers necessary to achieve a scaled score of 55 represents the cut score established by AICP as the minimum acceptable competency level. It is also important to note that scaled scores are not percentage scores. Scaled scores allow us to report different raw scores that represent the same kind of knowledge, skills, and abilities. These raw scores are converted, and all of the raw scores are mapped to the same or 'common' scale to produce the 'scaled scores.'

"Furthermore, the content area scores are not used to determine whether you passed or failed the test. They are included in the score report to indicate your strengths and weaknesses within major content areas and are provided for self-evaluation only. The total score is used to determine a passing grade, not content area scores.

"The Standard Error of Measurement (SEM) is one means of assessing the precision of a test score, which is never perfect. If you were to take many different forms or versions of the test on different occasions, your score would be different on each form but would cluster around a hypothetical 'true score.' This "true score" cannot be established absolutely because any test score can be influenced by a variety of factors: the condition of the candidate, the kind of test given, and other factors that are unrelated to the test itself. A candidate might perform differently on one occasion than on another. You might try harder, be more tired or anxious, have greater familiarity with the content of the questions on one test form than on another, or you might simply guess correctly more often on one test than another. All of these factors are taken into consideration when the SEM is calculated for a specific form of a test.

"The AICP test is 'equated' from one year to the next. Through the process of equating, we can make mathematical adjustments to scores so that one test form is comparable to another form of the test. . . . [For example,] through equating, we can determine that a score of 35 on one test represents the same level of math knowledge, skills, and abilities as a score of 40 on [another] test.