| Session Title: Utilizing Hierarchical Linear Modeling in Evaluation: Concepts and Applications |
| Demonstration Session 426 to be held in Calhoun Room on Thursday, November 8, 3:35 PM to 5:05 PM |
| Sponsored by the Quantitative Methods: Theory and Design TIG |
| Presenter(s): |
| Christopher McKinney, University of Northern Colorado, christopher.mckinney@mhcd.org |
| Antonio Olmos, Mental Health Center of Denver, antonio.olmos@mhcd.org |
| Kate DeRoche, University of Northern Colorado, kathryn.deroche@mhcd.org |
| Abstract: Hierarchical Liner Modeling (HLM) has become increasing popular in social and behavioral science fields. Evaluators who perform research within an organization and across organizations find this technique especially useful. The presentation will guide the participants through the major concepts of HLM and it's associated strengths, such as unbalanced data and missing values. Utilizing an example from the mental health field the processes of preparing data, running the model, model fit, and interpretation of the output will be discussed. A comparison to more common models, such a multiple linear regression and ANOVA will be utilized to demonstrate the effectiveness of HLM in the estimation of effects. Lastly discussion on other model types such as multinomial and logistic regression will be discussed to provide participants an understanding of HLM's flexibility. Throughout the presentation emphasis will be placed on interpreting and reporting HLM concepts and results to stakeholders and other interested parties. |