https://bjorn-persson.github.io/visualizations/Recent content in Visualizations on Björn PerssonHugo -- 0.147.2enhttps://bjorn-persson.github.io/visualizations/cfa/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/cfa/Interactive visualization of confirmatory factor analysis: see how factor loadings determine the model-implied covariance matrix, and watch an optimizer converge to the best-fitting parameter values.https://bjorn-persson.github.io/visualizations/construct-validity/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/construct-validity/Interactive comparison of six frameworks for construct validity: Cronbach & Meehl's nomological network, Campbell & Fiske's MTMM matrix, Messick's unified framework, Borsboom's causal realism, Kane's argument-based approach, and network psychometrics.https://bjorn-persson.github.io/visualizations/correlation-stability/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/correlation-stability/Compare five association measures interactively, then watch how sample correlations converge to the population value as N increases — based on Schönbrodt & Perugini (2013).https://bjorn-persson.github.io/visualizations/cronbachs-alpha/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/cronbachs-alpha/Interactive visualization of how Cronbach's alpha determines the Standard Error of Measurement and true-score confidence intervals — with direct consequences for psychological assessment.https://bjorn-persson.github.io/visualizations/irt-curve/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/irt-curve/Interactive visualization of Item Response Theory: explore how item characteristic curves, discrimination, difficulty, and information functions reveal measurement precision that Classical Test Theory hides.https://bjorn-persson.github.io/visualizations/irt-precision/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/irt-precision/Interactive visualization of how Item Response Theory determines measurement precision. Unlike Classical Test Theory, IRT precision varies across the trait continuum — explore how item parameters control where a test is most informative.https://bjorn-persson.github.io/visualizations/measurement-error/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/measurement-error/Interactive demonstration of how imperfect reliability attenuates observed correlations, and what the correction for attenuation costs in sampling variance. Adjust the true correlation and the reliability of each observed variable.https://bjorn-persson.github.io/visualizations/multicollinearity/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/multicollinearity/Interactive visualization showing how increasing the correlation between two predictors collapses the data cloud, inflates standard errors, and destabilizes regression coefficients. A 2D projection illustrates the geometric intuition.https://bjorn-persson.github.io/visualizations/normal-distribution/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/normal-distribution/Interactive visualization of the normal density extended with skew and kurtosis. Adjust the four moments and read off lower-tail, upper-tail, two-tailed, central, or range probabilities — or invert from a target probability.https://bjorn-persson.github.io/visualizations/tabulated-traits/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/tabulated-traits/Comparative overview of major perspectives on personality — behavior genetics, CB5T, Five-Factor Theory, Eysenck, Cloninger, evolutionary, lexical Big Five, HEXACO, interpersonal circumplex, socioanalytic, attachment, whole-trait, social-cognitive, psychodynamic, narrative, personology, humanistic/SDT, and social-constructionist — sorted along a reductionist–holistic continuum.https://bjorn-persson.github.io/visualizations/tabulated-psychiatry/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/tabulated-psychiatry/Comparative overview of the major perspectives on psychiatric disorders — biomedical, RDoC, computational, phenomenological, evolutionary, HiTOP, network, psychodynamic, biopsychosocial, cognitive, interpersonal, social/cultural, PTMF, humanistic, and critical psychiatry — sorted along an ontological realism continuum.https://bjorn-persson.github.io/visualizations/range-restriction/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/range-restriction/See how restricting the range of a variable attenuates correlations and reduces statistical power.https://bjorn-persson.github.io/visualizations/regression/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/regression/Interactive visualization for fitting linear, quadratic, cubic, and logistic regression models. Set population parameters, generate synthetic data, and compare the fitted curve to the true population curve.https://bjorn-persson.github.io/visualizations/regression-to-the-mean/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/regression-to-the-mean/Interactive scatterplot demonstrating how extreme scorers at Time 1 tend to score closer to the mean at Time 2 whenever the test-retest correlation is less than perfect. A reliability slider controls that correlation, and you can click to highlight top and bottom performers.https://bjorn-persson.github.io/visualizations/reliability-validity/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/reliability-validity/Interactive archery-target metaphor for Classical Test Theory. Two sliders control random error (reliability) and systematic error (validity), updating both the target cluster and the observed-score distribution.https://bjorn-persson.github.io/visualizations/simpsons-paradox/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/simpsons-paradox/Interactive scatterplot demonstrating how an overall negative correlation can reverse into positive within-group correlations when data is grouped. Drag the grouping slider to watch the paradox unfold.https://bjorn-persson.github.io/visualizations/dichotomization/Mon, 01 Jan 0001 00:00:00 +0000https://bjorn-persson.github.io/visualizations/dichotomization/See how splitting a continuous variable into two groups discards information, attenuates effect sizes, and reduces statistical power.