A growing body of recent research has shown that unmotivated test taking materially affects test-taker performance and degrades the validity of test score-based inferences.  This has led to a consequent need for methods of identifying and effectively managing unmotivated test taking.  Such methods require valid ways of measuring motivation, and one particular method, based on item response time, is unique in that it provides motivation information down to the level of the individual item responses.  When taking multiple-choice items, unmotivated examinees have been found to exhibit very short response times, which has been termed rapid-guessing behavior (as opposed to the solution behavior that is expected under the item response model).  Identification of rapid guessing is central to the indices of response time effort (which measures test taker effort) and response time fidelity (which measures how much effort individual items receive), as well as several other test delivery and scoring methods developed to mitigate the score distortion caused by unmotivated test taking.

This paper provides a formal examination of the theory, assumptions, and principles underlying the identification and interpretation of rapid-guessing behavior.  Key principles and methods are illustrated using data from a large-scale adaptive achievement testing program.  The issues discussed include:

• The theory of test taking behavior underlying by the rapid guessing model.
• Evidence regarding the model’s basic assumption that rapid guesses are uninformative about test taker achievement level.
• Evaluation of the assumption that rapid guesses are random, and its implications for their identification.
• A comparison of the strengths and limitations of various methods that have been proposed for identifying the response time thresholds used to classify responses as rapid-guessing behavior.
• An extended discussion of the normative threshold method as a recommended method for identifying rapid guessing.