From Screening to Intervention: Instructional Planning for Students Who Struggle with Whole Number Computation

Abstract

This dissertation conducted two studies that examined two methods of instructional planning to effectively match students to a whole number computation intervention that would best meet their needs. Study 1 was a systematic synthesis of all studies that used brief experimental analysis (BEA) to determine an effective mathematics intervention for students. Sixteen studies that included 67 participants and used a BEA to identify the most effective mathematics intervention were located. Results of Study 1 indicated that the majority of BEAs compared skill and performance interventions on computational fluency; however, the methodology across the included studies greatly varied. The second study evaluated a gated screening framework that included STAR Math, AIMSwebTM Math Computation (MCOMP), and a can’t do/won’t do assessment using AIMSwebTM Subskill Mastery Measure-Addition/Subtraction (SSMM-Add/Sub). A standard BEA was used to evaluate which of two interventions, modeling with error correction or explicit timing with reward, was most effective for each student. Analyses determined whether each of the screening measures could accurately differentiate between the students who benefitted the most from each intervention and accurately predict the outcomes of the BEA. Statistically significant differences were yielded for SSMM-rate but not STAR Math or MCOMP. STAR Math and SSMM-rate were able to predict which intervention was most effective. A cut score analysis indicated that the optimal cut score for SSMM-rate to differentiate between interventions was 13 DCPM.