Davood Tofighi, Ph.D., is a candidate for the Measurement, Evaluation, and Research Methodology (MERM) assistant professor (tenure-track) position.
Wednesday, April 29, 2015 | No RSVP required
| Teaching Talk | 10:00 – 10:45 a.m. | Neville Scarfe, Room 310 | “An introduction to hypothesis testing” |
| Colloquium + Q&A | 11:00 – 12:30 p.m. | Neville Scarfe, Room 310 | See abstract below. |
Abstract
Mediational analysis is a statistical approach that examines the effect of an antecedent variable (e.g., academic achievement) on an outcome (e.g., career aspiration in science) achieved by targeting and changing one or more intervening variable(s) (e.g., academic self-concept in science). Multilevel mediation analysis examines an indirect effect using multilevel (clustered) data such as students that are nested within classrooms. I first provide an overview of mediation analysis using a single mediator model and discuss the importance of mediation analysis in educational settings. I will then discuss my current research on the development and assessment of methods to test more complex mediation models such as a micromediational model with multiple mediators. Next, I will present the results of my recent study quantifying and diagnosing the biasing effects of potential omitted variables in a multilevel mediation model where the antecedent, mediator, and outcome are all measured at Level 1 (student level). I will use this to introduce how potential unmeasured variables can change the interpretation of the indirect effects and bias estimates of indirect effects at both student and school level. Finally, I will discuss remedies of how to assess the amount of bias by conducting a sensitivity analysis.
About Dr. Tofighi
Davood Tofighi is currently an assistant professor in the School of Psychology at the Georgia Institute of Technology. He received his Ph.D. in Quantitative Psychology from Arizona State University in 2010. Dr. Tofighi’s research interests include the development of novel statistical methodologies and software packages with applications in complex educational, psychological, and health research settings. His specific methodological interests are to develop robust tests with sufficient power and improve the accuracy and validity of inferential procedures in multilevel and longitudinal mediation analysis using both frequentist and Bayesian frameworks. His work has been published in journals such as Psychological Methods and the British Journal of Mathematical and Statistical Psychology. Dr. Tofighi received an M.A. in Educational Psychology and M.Sc. in Industrial and System Engineering from the University of Nebraska. He also obtained a B.Sc. in Industrial Engineering from Sharif University of Technology in Iran.