Computer-Based Testing (CBT)

NCEES Exams Computer-Based Testing

NCEES began the process of transitioning exams to computer-based testing (CBT) in 2011. CBT offers many benefits, such as enhanced security for exam content and more uniformity in testing conditions. For most exams, it also provides greater scheduling flexibility.

All CBT exams are offered at approved Pearson VUE test centers.

Man and Woman in hard hart going over blueprints and build

Reasonable accommodations are available for examinees who meet certain eligibility criteria and sufficiently document their request.

Alternative Item Types

CBT exams include traditional multiple-choice questions as well as alternative item types (AITs). AITs provide opportunities to assess the technical knowledge of examinees.

All questions, including AITs, are scored as either correct or incorrect. There is no partial credit. AITs include but are not limited to the following:

  • Multiple correct—allow examinees to select multiple answers
  • Point and click—require examinees to click on part of a graphic to answer
  • Drag and drop—require examinees to click on and drag items to match, sort, rank, or label
  • Fill in the blank—provide a space for examinees to enter a response to the question

Year-Round and Single-Day Exams

Some CBT exams are administered year-round. NCEES constructs these exams using a linear-on-the-fly (LOFT) algorithm. This means that all examinees for a particular exam are required to answer the same number of questions in the same topics; however, no examinees will have the same set of questions. The algorithm will assemble a unique exam within the same specification framework (i.e., the same number of questions per topic area) and the same relative level of difficulty.

Other CBT exams that have a smaller examinee population use a different high-stakes testing model and are administered on a single day each year. All examinees taking these exams receive the same questions.

The question formats used on both types of exams are the same, independent of the statistical model employed.