In OnTarget, when we talk about question reliability combined by subject, we’re referring to how consistent and dependable the test questions are when grouped by subject area (like Math, Reading, Science, etc.).
🔍 What is “question reliability”?
Question reliability refers to how consistently a set of questions measures student performance. A reliable test means that if students took a similar test again, they’d get similar results.
✅ What does “combined by subject” mean?
It means OnTarget is looking at all questions from the same subject area across one or more assessments, and analyzing them together to calculate a reliability score for that subject.
For example:
- If you’re analyzing Math, OnDataSuite will take all the Math-related questions from an assessment (or multiple assessments), and evaluate how well those questions work together to measure student performance.
- It does the same for Reading, Science, etc.
📊 How is reliability calculated?
OnTarget often uses statistics like:
- Cronbach’s Alpha – Measures internal consistency. Higher values (closer to 1) mean the questions are more reliable.
- Point-biserial correlation – Measures how well each question differentiates between high and low performers.