
Computational modeling provides a way to translate theory into explicit representations that can be simulated and examined, but many researchers lack a clear entry point for doing so. In this article, I introduce a two-step procedure designed especially for psychologists, particularly psychologists with little or no programming experience, who are interested in exploring computational modeling but are unsure how to start. The procedure involves two steps: (a) converting box-and-arrow diagrams (which many psychologists are already familiar with) into stock-and-flow (SF) diagrams and then (b) developing pseudocode from the SF diagram. Once these steps are completed, translating the model into executable syntax becomes straightforward. The goal is not to teach programming or simulation in depth but to show how conceptual models can be expressed in a form precise enough to generate trajectories over time. I then describe why this translation matters for theory. Explicit computational representations allow researchers to compare alternative explanations, evaluate whether simpler mechanisms are sufficient to reproduce observed patterns, examine whether theories generate the outcomes they claim, and explore ideas or policies that would be difficult to test directly in the field. Ready-to-use R code is provided to support hands-on learning to reach the broader aim of making computational modeling a practical and approachable tool for psychologists interested in evaluating their theories with computational modeling.
Advances in methodology and statistical software have made modern techniques for handling missing data more accessible and easier to implement. However, optimally applying these methods in moderation analysis remains challenging given assumption violations that arise from the nonnormality of interaction term when traditional approaches are used and nuances faced when modeling interaction terms in the presence of item-level missing data that must be reconciled with scale-level analyses. To address these issues, we present a comprehensive and practical user’s guide on implementing an advanced approach, factored regression specification (FRS), that ideally handles missing data in observed-variable-moderation models. Specifically, we provide step-by-step guidance for conducting FRS at both scale and item levels in various interaction scenarios that include missing data with both continuous and categorical variables. We also derive and present a new coding method to address the challenge of probing simple slopes in continuous-variable interactions that contain item-level missingness. Finally, we supply R code to facilitate implementation of the methods described and offer practical recommendations for researchers.
Cognitive neuroscience has long sought to disentangle stimulus-driven processing from conscious perceptual processing. Some prior evidence for neural processing of perceived musical beat (periodic pulse) may be confounded by stimulus-driven activity. Nozaradan et al. controlled for stimulus factors and used frequency tagging to show increased brain activity at imagery-related frequencies when listeners imagined a beat pattern during an isochronous stimulus. However, it remains unclear whether this effect is replicable and whether it reliably reflects conscious beat perception. In this registered report, we present 13 independent replications using the same vetted protocol. Listeners performed the same experimental paradigm as in Nozaradan et al. with an added behavioral task on each trial to assess conscious perception of the imagined beat. Preregistered meta-analyses revealed smaller raw effect sizes of imagery condition (binary = 0.03 µV, ternary = 0.03 µV) than the original study (binary = 0.12 µV, ternary = 0.20 µV) with confidence intervals all overlapping with zero. Differences in full-sample estimated effect sizes (this study: N = 152, η p 2 s = .03–.04; 2011 study: N = 8, η p 2 s = .62–.76) suggest larger sample sizes are necessary to detect these effects reliably if they exist. In addition, only neural activity at the stimulus frequency predicted imagery-task accuracy, contradicting our hypothesis that beat-related frequencies would predict performance. Our findings suggest an overall failure to replicate all main effects from the original study. We discuss potential reasons for discrepancies with the original study and implications for the utility of frequency tagging for studying beat perception.
Network analysis has become a popular computational-modeling approach in psychological research and is now increasingly used to address confirmatory research questions. However, preregistration of such analyses remains rare and even when implemented may leave room for researcher degrees of freedom, particularly when data-dependent decisions are difficult to fully anticipate. In this tutorial, we demonstrate how researchers can safeguard against bias by using analysis blinding as a complementary step to preregistration. The method involves temporarily altering the data to remove the key effect of interest while preserving all other aspects. Analysis blinding makes it possible to explore important features of the data, for instance, spotting outliers or adjusting the computational model, without introducing bias. In this tutorial, we present three common research questions in network analysis and show how they can be addressed using analysis blinding to safeguard their confirmatory nature. We use data from previously published studies employing network analysis and offer a practical, step-by-step guide along with R code that researchers can adapt to their own data and research questions.