UvA-DARE ( Digital Academic Repository ) Simpson ' s paradox in psychological science : a practical guide

By Rogier A. Kievit, Willem E. Frankenhuis, Lourens J. Waldorp, Denny Borsboom. 2017

Subfields: Statistics and Probability, Cognitive Science & Psychology

How researchers see this paper

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

In Statistics and Probability

In Cognitive Science & Psychology

Citations in the corpus

A Survey on Bias and Fairness in Machine Learning (2019, methodology): “Simpson’s paradox has been observed in a variety of domains, including biology [34], psychology [68], astronomy [89], and computational social science [77].”
But You Don’t Look Like A Scientist!: Women Scientists with Feminine Appearance are Deemed Less Likely to be Scientists (2016, unknown): “A good deal of research neglects to examine divergent sources of variance, which can mask important relationships in the data (Bell and Jones 2015; Kievit et al. 2013).”
Parental monitoring and adolescent problem behaviors (2016, background): “However, the paradox may arise with any inference that is drawn across different levels of observation, including from populations to subgroups, subgroups to individuals, but also from populations to intra-individual changes over time (Kievit et al., 2013).”
Repeated Measures Correlation (2017, background): “…(note in particular Panel (A), Row 1, and Panel (C), Row 3), in which patterns at a higher level of analysis (e.g., sample, experiment, study, or aggregated data) conflict with patterns at a lower level of analysis (Tu et al., 2008; Robinson, 2009; see Kievit et al., 2013; e.g., individual).”
Lack of group-to-individual generalizability is a threat to human subjects research (2018, background): “(39), who stressed that data should be explicitly tested for agreement in estimates across levels of analysis.”
Aligning AI Optimization to Community Well-Being (2020, background): “Attention to aggregates may also miss local problems, such as negative effects in a particular city or for a particular subgroup, or run into Simpsons ’ paradox issues where the sign of the effect depends on the granularity of the groups studied (Kievit et al. 2013).”
A critique of the cross-lagged panel model. (2015, background): “Furthermore, the correlation at the between-person level also influences the results, especially if it is of the opposite sign of the correlation that exists at the within-person level (i.e., in the presence of Simpson’s paradox, Kievit et al., 2013).”
A model for personality at three levels (2015, background): “A very clear exposition of the problem is found in Kievit et al. (2013).”
Mental disorders as networks of problems: a review of recent insights (2016, unknown): “Third, an important point is the generalization of group-level results to the individual level, since many group-level network studies have implied that the identiﬁed network structure of the population is more or less reﬂective of the networks of all individual participants (e.g., [68, 69]).”
Detecting Simpson's Paradox (2018, background): “to outcome analyses where both genders show negative trends, but gender differences drive a populaiton-wide positive trend (Kievit et al., 2013).”