Faculty coordinator: Dr. Paul DeBoeck
Venue: 35 Psychology Building
Time: 12:30-1:30pm
August 28, 2017
Welcome and introduction
September 4, 2017
Labor Day
September 11, 2017
Discussion of “Redefine statistical significance”
September 18, 2017
Speaker: Leanne Stanley
September 25, 2017
Speaker: Jack DiTrapani
October 2
in honor of Ragnar Steingrimsson
October 9
Speaker: Amanda Montoya
October 16
Speaker: Dr. Jolynn Pek
Department of Psychology, The Ohio State University
Title: On the utility of data transformations: Misuses, uses, and pedagogy
Abstract: Data transformations are a popular and easy-to-use tool in many researchers’ statistical toolkit. Families of transformations have been developed by eminent statisticians (e.g., Box & Cox, 1964; Tukey, 1977), and “street knowledge” seems to tout transformations as the quick fix to address data non-normality. This talk examines underlying motivations, repercussions on inferential devices (i.e., null hypothesis significance testing and confidence intervals), and the pedagogy of data transformations in psychological science. A nuanced appreciation of data transformations is pertinent to formulating theories, analyzing data, and accurately representing complex psychological data.
October 23
Speaker: Dr. Jessica K. Flake
Department of Psychology, York University
Title: The fundamental role of measurement quality in original and replicated research
Abstract: We are amidst an ongoing debate about the existence of a replication crisis, which has prompted scrutiny of our discipline’s practices. Registered reports, multi-lab replication efforts, and proposals for adjusting the statistical significance criterion level are a mere sampling of the unfolding shifts in our research culture. Although these proposals and discussions aim to increase transparency, rigor, and the quality of our work, they largely preclude the role of measurement practices. Despite that, measurement plays a fundamental role in the quality and replicability of our research. I will share the results of two systematic reviews of measurement practices in applied areas of psychology. In the first review of construct validity evidence, reported in a representative sample of articles published recently in The Journal of Personality and Social Psychology, I will demonstrate the frequent neglect of basic reliability and validity evidence of scales. From this review, I will highlight common, questionable practices which hinder the quality and replicability of our research. In the second part of my presentation, I will share preliminary results of a systematic review of the measurement quality of scales used in the original and replicated studies from the Reproducibility Project: Psychology (RPP). Across the studies of the RPP, scales from original research lacked adequate validity evidence and replicators commonly faced measurement challenges. I will expand on these challenges and discuss a framework for replicated research which explicitly incorporates rigorous measurement practices and how incorporating those practices will increase the value and impact of a replication.
October 30
Speaker: Dr. Brenden Bishop
Columbus Collaboratory
November 6
November 13
Speaker: Dr. Robert Cudeck
Department of Psychology, The Ohio State University
November 20
Speaker: Dr. Michael DeKay
Department of Psychology, The Ohio State University
November 27
Speaker: Dr. Yang Liu
Department of Human Development and Quantitative Methodology
University of Maryland, College Park
Title: Restricted recalibrations of item response theory models
Abstract: In item response theory (IRT), it is often necessary to perform a restricted recalibration
(RR) of the model: A set of (focal) parameters is estimated holding a set of
(nuisance) parameters fixed. Typical applications of RR include expanding an
existing item bank, linking multiple test forms, and associating constructs measured
by separately calibrated tests. In the current work, we provide full statistical theory
for the RR of IRT models under the framework of pseudo-maximum likelihood
estimation. We describe the standard error calculation for the focal parameters, the
goodness-of-fit assessment of the overall model, and the identification of misfitting
items. We report a small simulation study to investigate the performance of these
methods, which concerns adding a new item to an existing test. Parameter recovery
for the focal parameters as well as Type I error and power of the proposed tests are
examined.
December 4
Brief presentations on external talk
*pizza and soda will be served
Yiyang Chen
Seo Wook Choi
Bob Gore
Joonsuk Park
Saemi Park