Presentation on theme: "Dan Wright Sample City, KS. Common waiver component for several Kansas districts. Common Core & Smarter Balance emphasis on college/career readiness."— Presentation transcript:
Dan Wright Sample City, KS
Common waiver component for several Kansas districts. Common Core & Smarter Balance emphasis on college/career readiness. Availability of College Readiness Benchmarks in ACT series assessments.
Scaled scores in each subject area demonstrated to be associated with a 50% chance of getting a B or higher, or a 75% chance of getting a C or higher, in a credit- bearing freshman college course. How do these benchmarks compare with KCA performance?
Concurrent testing, Spring th Grade KCA General Ed, only (no KAMM) Universal Explore testing 1,150+ cases
KCA Reading Distribution at 8 th Grade
Explore Reading Distribution at 8 th Grade (Upper end of Explore distribution is only 25)
The joint KCA-Explore Reading distribution, shaped by both measures.
A number of approaches might be attempted to find a best fit. Linear regression was chosen, pending additional data. r =.58
College Readiness Benchmark KCA cut points and the Explore benchmark were applied. Joint Distribution of KCA and Explore Reading Exemplary Exceeds Standard Meets Standard
Explore Scale Scores Corresponding to Lower Ranges of KCA Performance Categories Explore READINGExplore MATH KCA Meets Standard 1213 KCA Exceeds Standard 1416 KCA Exemplary 1517 ACT College Readiness Benchmarks 1517
Concurrent Testing, Spring 2010 ACT cohort for 2011, tested in 2010 Spring, 2010 KCA Banked OTL from earlier years not used ACT-takers are self-selected Resulting group sizes of
Exemplary Exceeds Standard Meets Standard College Readiness Benchmark (Upper end of ACT distribution is 36) r =.63
ACT Scale Scores Corresponding to Lower Ranges of KCA Performance Categories ReadingMathScience KCA Meets Standard KCA Exceeds Standard KCA Exemplary ACT College Readiness Benchmarks
How to best fit the observed distributions. How to extend the sample into upper ranges. How might this change the predicted values. Implications for the next generation of assessments?