COMBATTING CEILING EFFECTS: MODELING HIGH-ABILITY STUDENT GROWTH USING MULTILEVEL TOBIT REGRESSION

Doctoral Candidate Name: 
Julia Hujar
Program: 
Educational Leadership
Abstract: 

Pressures associated with accountability testing have resulted in a narrowing of both the curriculum and pedagogy that does not meet the needs of high ability learners. This study proposed that either a different measurement (an above-level computer adaptive assessment) or a different model (Tobit model) should be used to more accurately demonstrate high ability student achievement and growth in order to lessen the pressures on teachers and therefore create an environment better suited for high ability student learning. To answer the research questions under study, a two-part design was used. The first part of the study used an above-level assessment and imposed an artificial ceiling at grade-level with the goal of using Tobit modeling to reproduce uncensored growth estimates using censored data. The second part of the study used naturally censored data with the goal of increasing growth estimates through Tobit modeling. Ultimately, the Tobit models using artificially censored data were able to come close to replicating the uncensored growth estimates under certain conditions. The results indicated that Tobit regression was necessary when examining homogeneous groups of high ability students. Finally, the Tobit regression models were able to increase the growth estimates for high ability students using naturally censored data. The degree to which the models increased, and under which conditions the increases existed are described in detail.

Defense Date and Time: 
Tuesday, May 24, 2022 - 2:30pm
Defense Location: 
Zoom
Committee Chair's Name: 
Richard Lambert
Committee Members: 
Michael Matthews, Kyle Cox, Stella Kim


The Graduate School and the Graduate Admissions office in the Reese Building, Fifth Floor, is temporarily closed to allow contractors to complete some needed work in the space safely.

Learn More