In this week’s episode Greg and Patrick talk about different ways of assessing inter-rater agreement and reliability among two or more raters and the importance of doing so. Along the way they also discuss the summer Olympics, underdogs, monologue face-offs, Quincy Wilson, Boomers, the Soviet judge, biopsy subjectivity, the secret to college admissions reliability, skipping […]
S6E02 Judging Inter-Rater Reliability Read More »
In this week’s episode, Greg and Patrick talk about the challenges of combining confirmatory factor analysis and multilevel data, and the underappreciated but absolutely critical role that theory plays in choosing the proper model for your constructs. Along the way they also discuss learning in a second language, torn meniscuseseses, concert C trumpets, nosy neighbors,
S5E21 Multilevel Factor Analysis: But What Do The Factors Mean?! Read More »
In this week’s episode Greg and Patrick discuss the assessment of global vs. local model fit and they argue that although global measures of fit can be useful, carefully assessing local fit may be of much greater importance in practice. Along the way the also discuss cheap beach house rentals, misplaced sand dunes, Mrs. Lincoln,
S5E20 Local fit…Because Global Fit Measures Suck! Read More »
In this week’s episode Patrick and Greg provide an introduction to the Item Response Theory model: what it is, how it relates to traditional factor analysis, and how this modem approach improves upon some of the limitations of classical test theory. Along the way they also mention weinerness, memorizing Latin for punishment, eggszampke, in ether
S5E19 Item Response Theory, Q.E.D. Read More »
In this week’s episode Greg and Patrick take a walk down memory lane to rediscover classical test theory, although they revisit this through the lens of modern latent variable models. They describe how these classical methods are actually highly restricted latent measurement models and they explore how these restrictions can be relaxed and even tested.
S5E15 Classical Test Theory (in bed) Read More »
In today’s episode, Greg and Patrick dig into Confirmatory Composite Analysis, a very clever way to get formative factors and their causal indicators into the traditional structural equation modeling framework, along with any other latent factors and their effect indicators that might already be in the model. Along the way they also discuss full-contact Wordle,
S5E08 Confirmatory Composite Analysis: Enter the Hexagon Read More »
In today’s episode, Patrick and Greg talk about the challenge of having causal indicators of formative factors within an analytical framework that is historically dominated by effect indicators and latent factors — and the critical importance of getting your arrows right. Along the way they also mention: self help books, habits, Hagrid and the giants,
S5E07 Perspectives on Formative Factors & Causal Indicators Read More »
In today’s episode, Patrick and Greg talk about fun extensions to the basic confirmatory factor model, including higher order models, bifactor or residualized models, and multitrait-multimethod models. Along the way they also discuss microscope lab, burning ants, substitute teaching, Cool or Creepy?, Monet, Clueless, haystacks, hotdogs, What are you thinking?, pennies and pounds, party like
S5E06 Advanced Factor Structures Read More »
In this week’s episode, the first for Season 5, Greg and Patrick explore the very cool world of receiver operating characteristic, or ROC, curves: what they are, how they work, and why we can give partial thanks to Winston Churchill for their existence. Along the way they also discuss advent calendars, lasagna for 8, honey
S5E01 Receiver Operating Characteristic (ROC) Curves Read More »
In this week’s episode Patrick and Greg talk about the critical and often unmet assumptions underlying the use of measured variables at multiple time points, whether for simple analyses like tests of means or more complex analyses like modeling individual growth. Along the way they also mention: following your blocker, Pascal’s wager, Wikipedia empathy, ham
S4E24 Longitudinal Measurement Invariance Read More »
