When a good fit can be bad.

By Mark A. Pitt, In Jae Myung. Trends in Cognitive Sciences, 2002

446 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/add2b3ff721d1366bda347f6d1e0c343d0ba7d7b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Comparing hemodynamic models with DCM (2007, unknown): “One also needs to account for differences in complexity; i.e., the number of free parameters and the functional form of the generative model (Pitt and Myung, 2002).”
Model Comparison and the Principle of Parsimony (2015, methodology): “The single most important quantitative element of model comparison relates to the ubiquitous tradeoff between parsimony and goodness-of-fit (Pitt & Myung, 2002).”
Detection of Differential Item Functioning Using the Lasso Approach (2015, background): “Because a particular -value balances fit (as indicated by the log likelihood) and complexity (indicated through the number of nonzero DIF parameters), the selection of an optimal parameter can be framed as a model selection problem that has desirable generalization properties (Pitt & Myung, 2002).”
Observing the Observer (II): Deciding When to Decide (2010, background): “Generally, pure model fit indices are not appropriate for comparing models and should be avoided (cf. [27-29]).”
Sensitivity and Specificity of Information Criteria (2018, unknown): “Many model selection techniques have been proposed in different settings (see, e.g., Miller, 2002; Pitt and Myung, 2002; Zucchini, 2000).”
Critical Thinking in Psychology: Evaluating Theories (2006, unknown): “One of the advantages of formal models of psychological phenomena is that they can be used to derive measures of how well a theory fits the data (Pitt & Myung 2002) that do not rely on null hypothesis significance tests.”
Bayesian Model Selection for Group Studies (2009, unknown): “Model comparison and selection is central to the scientific process, in that it allows one to evaluate different hypotheses about the way data are caused (Pitt and Myung, 2002).”
Prior sensitivity in theory testing: An apologia for the Bayes factor (2010, background): “At face value,many quantitativemeasures formodel evaluation seem to differ widely (see, e.g., Pitt & Myung, 2002; Shiffrin, Lee, Kim, & Wagenmakers, 2008, for overviews).”
The Importance of Falsification in Computational Cognitive Modeling. (2017, unknown): “, trading off goodness-of-fit with complexity) compared to rival models [8,12].”
Behavioral / Systems / Cognitive Striatal Prediction Error Modulates Cortical Coupling (2010, unknown): “Bayesian model selection When comparing different models for observed data, it is critical that the decision is not only based on the relative fit but also on the relative complexity of the competing models (Pitt and Myung, 2002).”
Nonlinear Dynamic Causal Models for fMRI (2008, unknown): “A decision about which of several competing models is optimal cannot be based only on the relative fit to the data but also needs to consider differences in model complexity; i.e., the number of free parameters and the functional form of the generative model (Pitt and Myung, 2002).”
Bayesian mixed-effects inference on classification performance in hierarchical data sets (2012, background): “In other words, the model evidence encodes how well a model strikes the balance between explaining the data and remaining simple (Pitt and Myung, 2002; Beal, 2003; Stephan et al., 2009).”
Ten Simple Rules for Dynamic Causal Modelling (2009, unknown): “Importantly, the model evidence is also a measure of generalizability (Pitt and Myung, 2002), i.e., how well the model generalizes across different data sets, and can thus be seen as an equivalent (but computationally more efficient) approach to procedures like crossvalidation (MacKay, 1992).”