Validating and Refining a Locally Developed Universal Screening System in High School Mathematics

Early identification of at-risk students is a critical element of Response to Intervention (RTI) but there are few validated universal screeners for high school mathematics. This study evaluated the effectiveness of the screening system currently in use at one urban Texas public high school and compared its effectiveness to alternate screening systems that used the same resources but with different cut scores. The goal for the screener was to accurately identify students who would Meet grade level proficiency on the state of Texas’ Algebra 1 End of Course exam. The current screening system used a multi-gating procedure that included the performance levels on the past state achievement test and on a beginning of year computer adaptive screening exam. Alternate systems were investigated using the assessments individually by exploring three different cut scores to determine if any were more effective. The study used the publisher’s cut score, the optimal cut score, and the cut score with a sensitivity of .94 (the same as the current system). The evaluation compared the overall accuracy, sensitivity, specificity, positive post-test probabilities, and negative post-test probabilities. No alternate system improved the effectiveness of the current system; however, none of the evaluated screening systems were able to adequately distinguish between at-risk and non-at-risk students. The threshold model is an identification model that excludes students with a low probability of failure from intervention, includes students with a high probability of failure in intervention, and requires additional information for those with no clear risk category. The use of a threshold model might result in a large number of unclassified students but could help identify those with the highest and lowest risk of failure. This study determined that a threshold model might support the development of a screening system in a school using an assessment with poor classification ability.

Grantor dc:publisher

University of Kansas

Year dc:date.issued

2021

Author dc:creator

• Stevens, Sarah E

Advisor dc:contributor.advisor

• Gay, A. Susan

• Educational evaluation
• Mathematics education
• Secondary education
• cut score
• evaluation
• RTI
• secondary mathematics
• threshold model
• universal screener

Statement dc:rights

• Copyright held by the author.

Language dc:language.iso

en

Dc Identifier Other

http://dissertations.umi.com/ku:17719

OAI identifier oai:identifier

oai:kuscholarworks.ku.edu:1808/39148