A scaled score transforms a raw test score (the number of test questions answered correctly) into other measurement units, called a scale score. However, please know that a scale score is not the number of questions answered correctly, nor the percentage of questions answered correctly.
Is the current test more or less difficult than prior tests, or future tests? There are multiple forms of the ICC examinations. While these forms were developed from the same set of content specifications, the levels of difficulty of the forms will vary because different exam questions appear on different forms. Some of these questions by their nature (and designated in the cut-score) are more difficult than others, even though they pertain to the same section of the exam. It would be unfair to require a candidate taking a collection (form) of somewhat more difficult questions to answer as many questions correctly as a candidate who took an easier form. So, we use a statistical procedure known as scaling to correct for differences in test form difficulty.
Think of it this way. Let’s say an elementary school hired the ICC to create an exam to test for knowledge of aspects of mathematics. Further, let’s say the exam was called “Addition and Subtraction,” and the Exam Catalog states it is a 100-item exam.
Look at the first ten questions of two different exam forms, both of which measure the concept of “Addition and Subtraction”:
|Question||Exam Form 1||Question||Exam Form 2|
|1.||2 + 4 =||1.||6 + 9 =|
|2.||6 + 4 =||2.||16 + 26 =|
|3.||7 – 3 =||3.||18 – 16 =|
|4.||18 + 15 =||4.||23 + 99 =|
|5.||16 – 4 =||5.||86 – 44 =|
|6.||8 + 8 =||6.||5 + 1 =|
|7.||4 + 9 =||7.||29 + 88 =|
|9.||4 + 15 =||9.||48 + 96 =|
|10.||7 – 4 =||10.||55 + 38 =|
As you can see, the questions on Form 2 are a bit harder than the questions on Form 1. If these were both 100-item exam forms with such a continuing mix of items, it would be unfair to require the students who were administered Exam Form 2 to answer the same number of questions correctly in order to pass.