DRC DRC Psychometrics and Research 2 — Questions and Answers
Question 1: What does 'vertical scaling' enable in DRC's multi-grade assessment program?
- Measuring how tall students have grown from year to year
- Placing scores from different grade levels on a common scale to measure growth across grades (Correct answer)
- Scaling test booklets to fit different physical page sizes
- Vertically aligning seating arrangements in testing rooms
Correct answer: Placing scores from different grade levels on a common scale to measure growth across grades
Vertical scaling creates a common score scale that spans multiple grade levels, enabling growth to be measured by tracking a student's score along the scale across years.
Question 2: What is a 'concordance study' in the context of DRC assessment research?
- A study examining agreement among test administrators
- Research establishing statistical relationships between scores on two different assessments (Correct answer)
- A study that concords (confirms) the accuracy of scores
- Research reconciling discrepancies in test shipping records
Correct answer: Research establishing statistical relationships between scores on two different assessments
A concordance study develops statistical crosswalk tables showing the relationship between scores on two different tests, helping users interpret and compare results across assessments.
Question 3: What does 'model fit' indicate when DRC applies IRT to test data?
- Whether the test booklet fits into the supplied envelope
- How well the chosen statistical model explains the actual pattern of student responses (Correct answer)
- Whether the test content fits the grade level curriculum
- How well the test scores fit within the expected normal distribution
Correct answer: How well the chosen statistical model explains the actual pattern of student responses
Model fit statistics evaluate whether the IRT model adequately describes the observed response data, with poor fit suggesting that model assumptions may not hold for certain items.
Question 4: What is 'test fairness' research as conducted by DRC psychometricians?
- Ensuring all students pay the same fees for testing
- Systematic investigation confirming that the assessment does not systematically disadvantage any student subgroup (Correct answer)
- Ensuring each student receives the same physical test materials
- Research into making test difficulty uniform for all schools
Correct answer: Systematic investigation confirming that the assessment does not systematically disadvantage any student subgroup
Test fairness research examines whether assessment processes, items, and scores produce comparable validity and reliability across diverse student subgroups.
Question 5: What is the purpose of 'linking studies' in DRC's multi-year assessment program?
- Linking test scores to student library card numbers
- Statistically connecting score scales across years so scores remain comparable over time (Correct answer)
- Linking multiple choice items to their corresponding constructed response items
- Connecting DRC assessments to other vendors' product lines
Correct answer: Statistically connecting score scales across years so scores remain comparable over time
Linking studies establish statistical relationships between test forms administered in different years, ensuring that a score of 250 this year means the same thing as a score of 250 last year.
Question 6: What does 'test information function' (TIF) show in IRT-based DRC assessments?
- A function that informs students about test content before testing
- A curve showing how precisely the test measures ability at different points on the ability scale (Correct answer)
- A list of information about each test form's development timeline
- The information provided to test coordinators about the testing program
Correct answer: A curve showing how precisely the test measures ability at different points on the ability scale
The test information function shows how much measurement precision a test provides at each level of student ability, with higher information indicating lower measurement error at that point.
What does 'vertical scaling' enable in DRC's multi-grade assessment program?