Definition
A method of evaluating individual test questions by examining how students performed on each one, using measures such as how many got it right, how well the question separated stronger students from weaker ones, and how often each wrong answer was chosen. The results show whether a question is doing its job or needs to be revised or replaced.
Plain English
Looking at the numbers from a test to see which questions worked well and which ones didn't, so the test can be improved.
Context Anchor
Seen in instructor training when building or reviewing written tests for aviation students.
Derivation
Statistical comes from statistics, meaning information collected and studied as numbers. Item, in testing, means one question on a test. Analysis means breaking something down to examine its parts. Together, the phrase points to studying each test question by using the numbers from student answers.
Why Pilots Care
A well-analyzed test measures what students actually know, rather than rewarding guessing or punishing students for poorly written questions. Instructors who use item analysis end up with fairer, more accurate exams.
Intuition Check
Do not read item as a physical object here. In this context, an item is one test question, and the analysis is about how that question performed when students answered it.
Example Sentence 1
After grading the written exam, the instructor ran a statistical item analysis and found that question 14 was answered incorrectly by nearly everyone, suggesting the question was poorly worded.
Example Sentence 2
Statistical item analysis showed one question had low discrimination, so it was rewritten before the next test.