nomologR - Guided Scale Development and Construct Validation
Provides an evidence-guided workflow for empirical scale development and construct validation, designed for graduate students and researchers. The package helps users examine item performance, dimensionality, measurement models, reliability, convergent and discriminant evidence, measurement invariance, and theory-specified nomological networks in the sense of Cronbach and Meehl (1955) <doi:10.1037/h0040957>. Methods are documented from historical to contemporary practice with literature references, and simulated teaching datasets with known population models accompany the examples. Following calls for transparent measurement practice such as Flake, Pek, and Hehman (2017) <doi:10.1177/1948550617693063>, it emphasizes transparent diagnostics, reproducible decision logging, and teaching-oriented guidance rather than automatic item deletion or binary declarations that a measure is valid.
Last updated
cfaconstruct-validityfactor-analysisfactor-scoreslavaanmeasurement-invariancepsychometricsreliabilityresearch-methodsscale-developmentstructural-equation-modeling
4.48 scoresolomonR - Analyze Solomon Four-Group Designs
Tools for planning, analyzing, teaching, and reporting Solomon four-group experiments, introduced by Solomon (1949) <doi:10.1037/h0062958>. The historical analyses are reproduced and labeled as such, from Solomon's improvement scores to the Tests A-I sequence of Walton Braver and Braver (1988) <doi:10.1037/0033-2909.104.1.150> and its published revisions. Contemporary analyses include one generalized linear model for all the groups, of four-group designs and of designs with several treatments, with robust or cluster-robust standard errors, for continuous, binary, and count outcomes; randomization inference; full-information maximum likelihood, following van Engelenburg (1999) <https://eric.ed.gov/?id=ED435692>; observed- and latent-variable structural equation models with 'lavaan'; and equivalence tests of pretest sensitization. Also includes design checks, power and sample-size planning, simulated teaching data, published data sets, and reports in the style of the American Psychological Association, with method-specific references.
Last updated
experimental-designpower-analysispretest-sensitizationrandomization-inferenceresearch-methodssimulation-studysolomon-four-groupstatisticsstructural-equation-modeling
4.08 score