Cutoff Scores, Cutting Scores
Cutoff scores, also referred to as cutting scores, are scores that differentiate the levels of performance.
The reliability of a measure is an essential factor to consider when using cutoff scores. Nunnally and Bernstein (1994) suggest that a test must have an internal consistency coefficient of at least 0.90, and ideally above 0.95, to designate a particular score as a diagnostic cutoff point, although this might be a high standard. For instance, if a child is demonstrating a learning disorder, the tests used to assess the disorder must be reliably able to measure the particular abilities, taking into consideration the span of the standard error, to place an education-altering label on that child. If such standards are not followed, then clinicians run the risk of basing a decision on a “false-positive” (the discovery of a limitation when none exists in actuality) or a “false-negative” (the missing of a limitation when one truly...
References and Readings
- Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill.Google Scholar