131 – Physical and Engineering Sciences
Novel Point Estimation from a Semiparametric Ratio Estimator (SPRE): Long-Term Health Outcomes from Short-Term Linear Data, with Application to Adults with Developmental Disability
Deborah Weissman-Miller
Brenau University
Catherine Link
The Shepherd Center
Celina Parkman
Cobb County School District
Point estimation is particularly important in predicting improvement in individuals who have developmental disabilities (DD). A new health response function based on a model of human response over time estimates long-term health outcomes from a change point in short-term linear regression in a pilot study for adults with DD. These estimations are given by parameters derived from short-term participant data in an (OLS) regression given in a new semiparametric ratio estimator model (SPRE). The response function is a ratio of two-parameter Weibull distributions times a prior outcome value stepping estimated outcomes forward in time. Shape and scale parameters are estimated at the change point. A feature of the SPRE model is that initial treatment response for each single-subject is reflected in long-term response to treatment. Results show that 7 participants had statistically significant results (p=0.05); a means analysis had statistically significant (p=0.00002) results showing an increase in skills from session 1 to session 12 (15.4%) and smaller increase for predictions (1.4%). Finally, the response function ratio provides a time frame for improvement for participants.