In this week’s episode Greg and Patrick discuss the sometimes terrifying issue of fungible weights in multiple regression and structural equation modeling in which selecting a trivially worse criterion of fit can often lead to radical changes in the corresponding parameter estimates. Along the way they also discuss competitive family Wordle, disambiguation, inflammability, perpitty, being nonplussed, running laps after practice, schmungible, audio eyerolls, Haystacks at Sunset, hyper eggs, the Spiderverse, mountain moonrises, tin cans and strings, and Earthquake Waller.
Lightly-Edited Episode Transcript
We provide a lightly-edited and imperfect audio transcript of the episode available here. This is not an exact representation of the audio, but does provide a searchable document with identified speakers and associated time stamps.
Additional Show Notes
Jones, J. A., & Waller, N. G. (2016). Fungible weights in logistic regression. Psychological Methods, 21, 241–260.
Lee, T., & MacCallum, R. C. (2015). Parameter influence in structural equation modeling. Structural Equation Modeling: A Multidisciplinary Journal, 22, 102-114.
Lee, T., MacCallum, R. C., & Browne, M. W. (2018). Fungible parameter estimates in structural equation modeling. Psychological Methods, 23, 58-75.
MacCallum, R. C., & Lee, T. (2012). Fungible parameter estimates in latent curve models. In Current topics in the theory and application of latent variable models (pp. 207-221). Routledge.
Pek, J., & Wu, H. (2015). Profile likelihood-based confidence intervals and regions for structural equation models. Psychometrika, 80, 1123–1145.
Pek, J., & Wu, H. (2018). Parameter uncertainty in structural equation models: Confidence sets and fungible estimates. Psychological Methods, 23, 635–653.
Prendez, J. Y., & Harring, J. R. (2019). Measuring parameter uncertainty by identifying fungible estimates in SEM. Structural Equation Modeling: A Multidisciplinary Journal, 26, 893-904.
Waller, N. G. (2008). Fungible weights in multiple regression. Psychometrika, 73, 691–703.
Waller, N. G., & Jones, J. A. (2009). Locating the extrema of fungible regression weights. Psychometrika, 74, 589–602.